Approaching the Viking Age: Proceedings of the International Conference on Old Norse Literature, Mythology, Culture, Social Life and Language, 11-13 October 2007, Vilnius, Lithuania 9955334924, 9789955334927

Citation preview

Approaching the Viking Age

2

ndina

 Vilnen

ic vist a  2

sis Sca

Vilniaus universitetas Skandinavistikos centras

Vikingų epochos tyrinėjimai Tarptautinės konferencijos „Dialogai su vikingų epocha“, vykusios 2007 m. spalio 11–13 d. Vilniuje, medžiaga Sudarė Erika Sausverdė ir Ieva Steponavičiūtė

Vilniaus universiteto leidykla · 2009

Centre of Scandinavian Studies Vilnius University

Approaching the Viking Age Proceedings of the international conference on Old Norse literature, mythology, culture, social life and language 11–13 October 2007, Vilnius, Lithuania Edited by Ērika Sausverde and Ieva Steponavičiūtė

Vilnius University Publishing House · 2009

UDK / UDC 839.6.09(06) Ol-15 Konferenciją parėmė / Conference sponsored by Šiaurės ministrų tarybos kultūros fondas / Nordic Culture Fund Lietuvos valstybinis mokslo ir studijų fondas / Lithuanian State Science and Studies Foundation Knygos leidimą parėmė / Publication sponsored by Šiaurės ministrų tarybos kultūros fondas / Nordic Culture Fund Serijos Scandinavistica Vilnensis redakcinė kolegija / Editorial board for Scandinavistica Vilnensis series Dr. habil. Jurij K. Kusmenko (Rusijos mokslų akademijos Kalbotyros institutas, Sankt Peterburgas, ir Humboldtų universitetas, Vokietija / Institute of Linguistic Studies under the Russian Academy of Sciences, Saint Petersburg, and Humboldt University, Germany) Dr. phil. Anatoly Liberman (Minesotos universitetas, JAV / University of Minnesota, USA) Dr. Ērika Sausverde (Vilniaus universitetas / Vilnius University) Dr. Ieva Steponavičiūtė (Vilniaus universitetas / Vilnius University) Dr. Aurelijus Vijūnas (Valstybinis Gaošiongo universitetas, Taivanas / National Kaohsiung Normal University, Taiwan) Recenzentai / Reviewed by Dr. Rasa Ruseckienė (Mokslo ir enciklopedijų leidybos institutas / Lithuanian Science and Encyclopaedia Publishing Institute) Dr. Aurelijus Vijūnas (Valstybinis Gaošiongo universitetas, Taivanas / National Kaohsiung Normal University, Taiwan) Apsvarstė ir rekomendavo spaudai Filologijos fakulteto taryba (2008 11 07, protokolas  Nr. 41) / Approved for publishing by the Council of the Faculty of Philology (07 11 2008, record No. 41) Knygos dailininkas / Designer Tomas Mrazauskas

© Vilniaus universitetas / Vilnius University, 2009 ISSN 2029-2112 ISBN 978-9955-33-492-7 Vilniaus universiteto leidykla / Vilnius University Publishing House Universiteto g. 1, LT-01122 Vilnius Tel. +370 5 268 7260 · www.leidykla.eu Vilniaus universitetas · Skandinavistikos centras Centre of Scandinavian Studies · Vilnius University Universiteto g. 5, LT-01513 Vilnius Tel. +370 5 268 7235 · www.skandinavistika.flf.vu.lt

Contents

The Editors’ Note  / 7 Evaldas Grigonis, Ērika Sausverde. Joachim Lelewel, Edda, and Lithuania  / 9 Rasa Baranauskienė. Reflections of Celtic Influence in Hildinavisen  / 15 Jan Ragnar Hagland. On Translating Icelandic Sagas into Modern Norwegian – the Case of Brennu Njáls Saga  / 41 Jon Gunnar Jørgensen. Norse Kings’ Sagas Spread to the World  / 55 Jurij K. Kusmenko. Sámi and Scandinavians in the Viking Age  / 65 Anatoly Liberman. Þjalfi  / 95 Ugnius Mikučionis. Norwegian Modal Verbs and Attitudinal Modality  / 117 Else Mundal. The View of Blood Vengeance in Medieval Norwegian Sources  / 139 Agneta Ney. Far och son-motiv i nordisk hjältediktning och saga  / 153 Ieva Steponavičiūtė. Saga Reflections in Karen Blixen’s Texts (with Focus On Grjotgard Ålvesøn og Aud)  / 163 Vésteinn Ólason. Scenes from Daily Life in the Sagas of Icelanders: Narrative Function  / 179 Kristel Zilmer. On Some Pragmatic and Symbolic Features of Island Representation in Old Norse Literature  / 197

The Editors’ Note

In autumn 2007, the Centre of Scandinavian Studies at Vilnius University hosted an international conference on Old Norse literature, mythology, culture, social life and language, held on the occasion of the 200th anniversary of the publication in Vilnius of Edda Skandinawska by Joachim Lelewel. The title of the conference − Dialogues with the Viking Age − was borrowed from the book by Vésteinn Ólason, with the author’s kind consent. The conference became a momentous event in the history of the Centre, as it gathered together many of our good friends from throughout the years: Professors Thomas Bredsdorff (Denmark), Jan Ragnar Hagland and Jon Gunnar Jørgensen (Norway), Jurij K. Kusmenko (Russia and Germany), Vésteinn Ólason (Iceland), Dr. Kristel Zilmer (Estonia) and Dr. Aurelijus Vijūnas (Taiwan). Although, for obvious reasons titled differently, the present book consists of the proceedings of that conference almost in their entirety, as well as a couple of extra articles by those authors who were not able to come to Lithuania at that time. We are grateful to all contributors, reviewers and sponsors who made it possible for the conference to take place and for the book to appear. Special thanks go to Ugnius Mikučionis and Dr. Aurelijus Vijūnas for their generous help in, respectively, organising the conference and editing the book. Ērika Sausverde Ieva Steponavičiūtė

Joachim Lelewel, Edda, and Lithuania Evaldas Grigonis, Ērika Sausverde Vilnius University

What is the connection between these three topics? 200 years ago, in 1807, Joachim Lelewel published a small book known as Edda Skandinawska in Vilnius. The book was written in Polish, and it was the first retold variant of Edda published in Lithuania. It is a significant event, since we have not been spoiled by numerous editions on or of Old Norse literature in Lithuania. Actually, only a few works of Old Norse literature have been translated into Lithuanian so far. Among them, there is Egil’s saga,1 The Saga of Knytlings,2 an anthology on Old Norse literature, which contains several excerpts of The Book of Icelanders, The Book of Settlement, The First Grammatical Treatise, several fragments from the Poetic and the Prose Edda, as well as excerpts from several other sagas.3 Therefore, every attempt to present this marvellous literature to the reader – the modern or the old days one – is worthy of attention and discussion. The precise name of Lelewel’s book is Edda czyli księga religii dawnych Skandynawii mięszkańców (‘Edda or the book on religion of the Old Scandinavian inhabitants’). It contains a chapter on the Old Germanic, Celtic and other tribes, a short presentation of both Eddas and finally a short paraphrase of the Eddas, first of all of the Younger Edda (only Völuspá and Hávamál from the Elder Edda are briefly discussed), supplied by an index of proper names. 1 Egilio saga. Iš senosios islandų kalbos vertė Svetlana Steponavičienė. Vilnius, Vaga, 1975. 2 Kniutlingų saga. Iš senosios islandų kalbos vertė Ugnius Mikučionis. Vilnius, Vaga, 2002. 3 Mimiro šaltinis. Senųjų islandų tekstų antologija. Sudarytoja ir vertėja Rasa Ruseckienė. Vilnius, Vilniaus universiteto leidykla, 2003. See also Skaldų poezija. Parengė Rasa Ruseckienė, Ugnius Mikučionis. Vilniaus universitetas, 2002. Approaching the Viking Age. Proceedings of the International Conference on Old Norse Literature, Mythology, Culture, Social Life and Language. Edited by Ērika Sausverde and Ieva Steponavičiūtė. (Scandinavistica Vilnensis 2). Vilnius University Press, 2009. https://doi.org/10.15388/ScandinavisticaVilnensis.2009.2.1 Copyright © 2019 Authors. Published by Vilnius University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Joachim Lelewel (1786–1861), the author of Edda Skandinawska.

One should keep in mind that Joachim Lelewel was 21 years old at the time of the publication of this book, and it appeared 11 years before his contemporary Rasmus Rask published the first complete edition of both Eddas in Stockholm in 1818, the latter event having evoked enormous interest in the Old Norse literary heritage in Europe. Though Lelewel’s personality has been widely studied and is wellknown, we would like to mention several details from his biography, which may be of interest in the context of our conference. First of all, he was a well-known, probably the greatest Polish historian of the first part of the 19th century, a scientist-erudite, bibliophile and polyglot. From his paternal side, he is of Austrian or Swedish ancestry,4 descended from early immigrants to Poland via Prussia and France, where Lelewel’s grandfather finally settled down in 1732.5 Wherever Lelewel’s ancestors came from, Lelewel showed himself as a very active Polish public character and

4 See Polski słownik biograficzny. Wrocław [etc.], 1972, t. 17/1, z. 72, s. 25. 5 See Piročkinas, A.; Šidlauskas, A. Mokslas senajame Vilniaus universitete. Vilnius, 1984, p. 270; cf. Śliwiński A. Joachim Lelewel: Zarys biograficzny. Lata 1786–1831. Warszawa, 1932, s. 3–4.

11  Joachim Lelewel, Edda, and Lithuania politician – all his life he fought and defended democratic and republican ideals. He was elected deputy of the Polish Parliament in 1829. He enthusiastically joined the uprising of 1830–1831, and was one of its most active participants. He was even a member of the rebel government, responsible for faith affairs and education. Appealing to Russian soldiers, he created the slogan “Za wolność naszą i waszą” (‘For our and your freedom’). Nevertheless, it was not only his political activities which made him a prominent figure in the Polish history. In his time, he amazed the academic society with his innovative conceptions and incredible diligence. Lelewel studied at Vilnius University (1804–1808, mostly classical philology under the guidance of Gottfried Ernest Groddeck), where later he started to teach history (1815–1818), and eventually became a professor (1822–1824). He must have been a brilliant lecturer, and his lectures on Polish history were enthusiastically accepted not only by academic youth, but also by a broader audience. His first lecture had to be postponed because of an overcrowded auditorium – around 1500 listeners came to listen to him! His further lectures used to draw crowds of about 400 people.6 His courses fitted a romantic world-view and had a big influence on students, especially Adam Mickiewicz and the Philomath Society (a conspiratorial student organization at Vilnius University in 1817–1823).7 In 1824, after the arrest and trial of the most active members of the society, Lelewel was expelled from University as a persona non grata for the Russian Tsarist regime. Afterwards he lived in Warsaw, and later in emigration in Paris and Brussels. As mentioned, his scientific productivity was imposing – up to 1830, he published about 150 research works. The scope of his interests was immense – from history and historical geography to numismatics, cartography and bibliography. From Edda Skandinawska, which is of special interest to us, to, for example, Géographie des Arabes (1851). Many of Lelewel’s books remain as classics in their fields. Edda was actually the first published treatise of the young Lelewel. This book, along with several other books authored by Lelewel is kept at the Rare Books Department of Vilnius University library. We would like to mention at least some of them.

6 Śliwiński, A., op. cit., s. 116–120. 7 See, e. g., Venclova, T. Vilniaus vardai. Vilnius, 2006, p. 122.

A facsimile of Edda Skandinawska, 1807

Historyka tudzież o łatwem i pozytecznem nauczaniu historyi, Wilno, 1815. In this work Lelewel raises a new viewpoint on the historical method, considering the problem of the importance of historical sources. Lelewel’s capital work on historical geography is Badania starożytności we względzie geografji, Wilno, Warszawa, 1818, widely accepted by leading European geographers of the 19th century, such as Alexander von Humboldt, Karl Ritter and others.8 Dzieje starożytne Indji…, Warszawa, 1820 was the first – and for a long time the only – work on Indian culture in Polish. Historiczna parallela Hiszpanii z Polską w wieku XVI, XVII, XVIII, Warszawa, 1831 was an innovative work, applying the comparative method in history. Numismatique du moyen-âge, Paris, 1835. This is the most significant of Lelewel’s work in numismatics, where he discusses and classifies Belgian, French, German, Czech, Swedish, Polish and other early medieval coins. Presumably, the publication laid the foundations of the medieval numismatics. Finally, one should mention the 5 volumes of La Géographie du moyen-âge, Bruxelles, 1849–1857.

8 Vilnius University professor Gottfried Ernest Groddeck commented on the publication, that even if Lelewel had not written anything else this work would have been enough. (Śliwiński, A., op. cit., s. 90).

This is the largest Lelewel’s work in historical geography and cartography, which remains classical up to now. The work was complemented by atlases with 177 maps, the plates entirely engraved by Lelewel himself (some of the maps even tinted in watercolors). Lelewel died in Paris in 1861. He was buried in Montmartre, but in 1929, his remains were moved to the Rasų cemetery in Vilnius. In his will, Lelewel left his library to Vilnius University. The will was implemented in 1926. The University got about 4800 books and 397 atlases and maps.9 Thanks to Lelewel, Vilnius University is proud to have one of the best atlas collections in Europe. Joachim Lelewel wrote in Edda, that if somebody would like to get to know more about this, let him read the Edda itself.10 With these Lelewel’s words, let us proceed to the Eddas and sagas themselves. 9 Braziūnienė, A. Senosios kartografijos rinkiniai Lietuvoje. Knygotyra, t. 36, 2000, p. 36–37. A part of the collection can be seen in the room named in Lelewels’s honor at Vilnius University library. 10 Lelewel, J. Edda czyli księga religii dawnych Skandynawii mięszkańców. Wilno, 1807, s. 54.

Reflections of Celtic Influence in Hildinavisen Rasa Baranauskienė Vilnius University

The Orkneys seem to have been of particular importance in transmitting certain Celtic material. Gísli Sigurðsson argues, however, that it is more likely that cultural contacts and exchanges which took place in the Orkneys between Icelandic and Gaelic-speaking people were limited to single motifs, tales or poems. This does not mean that the single features are limited in number, only that they are found as single items in a tradition which had to be built up in Iceland from the cultural elements available in the country itself (Gísli Sigurðsson 1988: 42). Though the Orkneys were an ideal meeting place where Scandinavian and Celtic cultures could exchange traditions, the Gaelic custom in question existed in Iceland as well, having been brought there by the Gaelic settlers. It is another matter that these traditions could be reinforced because of the contacts in the Orkneys (Gísli Sigurðsson 1988: 44). Among the most important elements identified as transmitted through the contacts via the Orkneys are stories including Hjaðningavíg ‘The Battle of Hjadnings’. Einar Ólafur Sveinsson assumes that Celtic tales played an important part in forming Icelandic ideas about the everlasting fight – a motif which becomes extremely common in Iceland, relevant right down to the 19th century, but is rare in the Scandinavian tradition (Einar Ólafur Sveinsson 1959: 17–18). Háttalykill, ‘Clavis metrica’ or ‘Key to Metres’ composed in the Orkneys in the 1140 by an Icelander and the Orkney Earl Rognvaldr kali, contains what is believed to be the earliest reference to the ‘Everlasting fight’ motif in Old Norse / Icelandic literature, the motif being taken over from the Irish 9th century tale Cath Maige Tuired ‘The Battle of Mag Tuired’ (Chesnutt 1968: 132). The literary works which are likely to have emerged from the cultural mixture in the Orkneys and the Scottish Isles are poems such as Darraðarljóð and Krákumál (Holtsmark 1939: 82). Konungs Skuggsjá contains a passage on Ireland which shows similarities to Topographia Approaching the Viking Age. Proceedings of the International Conference on Old Norse Literature, Mythology, Culture, Social Life and Language. Edited by Ērika Sausverde and Ieva Steponavičiūtė. (Scandinavistica Vilnensis 2). Vilnius University Press, 2009. https://doi.org/10.15388/ScandinavisticaVilnensis.2009.2.2 Copyright © 2019 Authors. Published by Vilnius University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

16  Rasa Baranauskienė Hibernica of Giraldus Cambrensis (Holtsmark 1964: 667). The source may have been a written one. Michael Chesnutt goes further and claims that Latin may have been the medium of communication (Chesnutt 1968: 135). Considering the transmission of various motifs, it seems that Orkney was a channel through which motifs could travel from Scandinavian to Celtic areas and vice versa. Celtic material, such as single motifs, tales or poems could also reach the surrounding areas, above all Shetland Islands. The Shetland material has never been discussed before in this particular aspect, though the Shetland Islands were clearly on the route for the transmission of the Celtic material. Ideas from the Orkneys could be easily transmitted to Shetland (the distance between the clusters of Orkney and Shetland Islands is not so long). The Norn ballad Hildinavisen to be discussed in this chapter, which seems to contain Celtic elements, has clearly been influenced by the Orkney tradition, because its main protagonist is Jarlin d’Orkneyjar ‘Earl of Orkney’. However, we should not imagine that the Orkney Earldom was always an intermediary in the exchange of such tales. There was also direct Gaelic influence on the Western Scandinavian, especially Icelandic and Faroese, folk tradition, since some of the Scandinavian settlers on these islands came via Ireland and Scotland and had sometimes lived there and absorbed Gaelic culture (Almqvist 1981: 89). Language of the Settlers The variety of Scandinavian language in Orkney and Shetland came to be called “Norn”. “Norn” is a contracted form of the feminine adjective norr n (from Old Norse norðr nn, meaning ‘coming from the north’), which in the 13th century came into use in order to differentiate between the western and eastern variety of Norse. Until then, all Scandinavian languages were called “Do˛nsk tunga”, a term which has survived in Orkney dialect till now (Rendboe 1987: 1). The term “Norn”, meaning ‘(Western) Norse language’, ‘Norwegian language’, ‘(Western) Norse’ or ‘Norwegian’ was first recorded in an endorsement in Scots appended to a Norwegian document from 1485 dealing with Shetland matters. Although it was also occasionally applied to Norse speech elsewhere in Scotland (Barnes 1996: 21), Norn is in most contexts used exclusively of the Northern-Isles variety (Barnes 2000: 179). Norn was not a dialect,

17  Reflections of Celtic Influence in Hildinavisen since its speakers appear to have considered that they have formed their own speech community in the same way as the speakers of Faroese (Barnes 1996: 13). These islands retained their Scandinavian character for a long time, even after they had been pledged to King James III of Scotland in 1468–9 (Barnes 2000: 173). The Sources in Norn The settlers and their descendants have not left so many written sources, either in runes or the Roman alphabet, and the few texts that exist tend to mirror faithfully the contemporary idiom of Norway (Barnes 2000: 179). There exist a few runic inscriptions from the 11th century, some old diplomas – the oldest one is from 1299 written in Old Norse, and some of them are written in Old Danish. Another coherent text is James Wallace’s Orkney version of the Lord’s Prayer (source unknown) published in the second edition of An Account of the Islands of Orkney, written in 1700 (Wallace 1700). Probably the most interesting sources have been registered by George Low from Edzell in Angus, but resident in Orkney – more an amateur student of natural history than a linguist. He recorded samples of Norn from the Shetland island of Foula during a visit there in 1774 from the 19th of June until the end of August and included them in his book A Tour through the Islands of Orkney and Shetland, first published in 1879. Low has also recorded a Shetlandic version of the Lord’s Prayer (source unknown), and a list of thirty English words translated into Norn, presumably by different informants. Another text is a 35-stanza ballad obtained from an old man “William Henry, a farmer in Guttorm, in Foula”. This old folk song, never registered before, is now popularly called Foulavisen or Hildinavisen, after the heroine. It has been assumed that the language of this song is much older and represents the language from the 1660 (Flom 1925: 127). I would argue that its language could be even older, because of its poetic form, which usually conserves the grammatical form and content. Other material comes from the period when Norn was no longer a living language. Here the most important are Jakob Jakobsen’s collections. This scholar from the Faroe Islands called Jákup doktari contributed significantly to the understanding of individual Scandinavian words (Barnes 1996: 1). His dictionary contains ca. 10 000 items. Hugh

18  Rasa Baranauskienė Marwick’s glossary of Norn contains ca. 3000 items. The scantiness of sources is frustrating, especially when it comes to coherent texts. The Norn Ballad Though the ballad Hildinavisen is recorded in Foula, it does not mean that it was composed there. There were plague epidemics in 1700, 1720 (when just six inhabitants out of ten survived), 1740, 1760 and 1769. So Hildinavisen could have been brought from the mainland of Shetland, because people were moving to the islands, especially to the islands further North. Hildinavisen was sung by William Henry, a farmer in Guttorm in Foula. Low writes about the old man who recited Hildinavisen in a letter from 1776, saying that he “could neither read nor write, but had the most retentive memory I ever heard of” (Low 1879: 107). In his book Low claims that “the following song is the most entire I could find, but the disorder of some of the stanzas will show that it is not wholly so… Here it is worthy to be observed that most of the fragments they have are old historical Ballads and Romances, this kind of poetry being more greedily swallowed and retentively preserved by memory than any others”. Low continues that “he [William Henry] spoke of three kinds of poetry used in Norn, and repeated or sung by the old men; the Ballad (or Romance, I suppose); the Vysie or Vyse, now commonly sung to dancers; and the simple Song. By the account he gave of the matter, the first seems to have been valued here chiefly for its subject, and was commonly repeated in winter by the fireside; the second seems to have been used in public meetings, now only sung to the dance; and the third at both” (Low 1879: 107). He also notes that William Henry “repeated and sung the whole day” (Hægstad 1900: 11). The most peculiar moment of recording is that Low did not know any Scandinavian language, i. e. he did not understand a word of what he was writing down. He writes: “In this Ballad I cannot answer for the orthography. I wrote it as an old man pronounced it; nor could he assist me in this particular” (Low 1879: 107). Having in mind the words of William Henry, one can call this ballad a diffuse continuum of an earlier epic tradition. The analysis of a ballad is a complicated matter, because the recording is fault. The late date of the recording is one factor. But at least we can be sure that when this ballad was recorded, Norn still was a living language. This is

19  Reflections of Celtic Influence in Hildinavisen indicated not just by Low, but also by other sources, such as The Description of the Isles of Orkney and Zetland, published in Edinburgh in 1771 by Sir Robert Sibbald. He describes conditions in Shetland around 1680 as such: “All the Natives … can speak the Gothick or Norwegian Language, and seldom speak other among themselves” (Sibbald 1771: 48–49). It seems though that this situation soon changed, and, according to Laurits Rendboe, Norn sang its swansong i. e. lived its last days (Rendboe 1987: 6). This ballad is the only complete ballad in Norn (there are no other complete ballads either from Orkney or from Shetland). Moreover, it was transmitted orally and Low presented what he heard through the medium of English, and to a limited extent French orthography. Thus we have some issues that complicate the analysis of the ballad: the lack of other similar texts in Norn, the incomplete information on which projections are made, as well as the paucity of texts. The fact that Low did not understand what he was writing may indicate that the text, as we have it, reflects its pronunciation. Besides, the bound form of the ballad ensures its longevity in an unchanged form. However, “stylistically, one would not expect … a (medieval?) ballad to reflect everyday speech” (Hammersheib 1981: 181). At the same time, it is to be expected that the language of this ballad was preserved by its bound form, and since the ballad was sung, the melody must have preserved the form and contents unchanged.1 This type of conservation allows us to think that the language of the ballad is more or less archaic and its content has been preserved without drastic changes. The Sources of Scandinavian Material There is a voluminous tradition related to the names of the two protagonists of the Shetlandic ballad – Hiluge and Hildina – in Scandina­vian sources. 1) First of all, it goes back to Illuga saga Gríðarfóstra ‘The Saga of Illugi, Grid’s Foster Son’ which belongs to the bulk of Fornaldarsögur Norðurlanda ‘Sagas of Ancient Times’. However, Illuga saga Gríðarfóstra is not preserved in any of the  collections of Fornaldarsögur from the Middle Ages. The oldest manuscript of this saga is AM 123 8vo, which

1 The fact that it was sung at the time when it was recorded is known from the letter of Low mentioned above.

20  Rasa Baranauskienė is on parchment, but hardly much older than 1600 (Davíð Erlingsson 1975: 11). With regard to Illuga saga Gríðarfóstra, Knut Liestøl supports the traditional opinion that it dates back to about 1300 (Liestøl 1958: 125). The writing of Fornaldarsögur is generally believed to have started at the end of the Golden Age of Icelandic literature in the late 13th century. It became increasingly popular in the 14th century, when most of these sagas are thought to have been written down (Sigurðsson 1988: 48). Secondly, there are quite a few ballads containing names similar to the protagonists’ of Hildinavisen: 2) Two versions of the ballad (A and B) found in the Faroe Islands are called Kappin Illugi. The second protagonist is Hilda (in the version A) or Hildur (in the version B). Both versions of the Faroese ballad were published by Venceslaus Ulricus Hammersheib in F röiske kv der, 2. 3) A version of a ballad which has a similar story is found in Norway and is called Kappen Illhugin and was published by Magnus Brostrup Landstad in Norske Folkeviser, nr. 2. 4) In Denmark, the ballad is called Herr Hylleland henter sin jomfru. It was published by Sven Grundtvig in Danmarks gamle Folkeviser, nr. 44. The difference is that here the male protagonist is called Hylleland instead of Illugi. 5) Finally, there is the Shetlandic ballad called Foulavisen or Hildi­ na­vi­sen which contains protagonists with the same names. However, the  content of the  Shetland ballad differs greatly from the  Faroe, Norwegian and Danish ballads, as well as from the Icelandic saga. It is of great importance that Hildinavisen contains completely different features and even a different story which accommodates the so-called “Celtic love triangle” and where the new character Jarlin d’Orkneyar is introduced. Other major aspects that make it specific are Hiluges negative character and the female protagonist’s name Hildina. The etymology of the name Hiluge (illr ‘ill’ + hugr ‘mind’) might indicate that it was attached to an evil personage or a troublemaker from the very beginning, similar to Bricriu and Efnisien in Celtic medieval literature. Celtic Motifs in Hildinavisen Comparative literature studies enable us to discover various types of connection between the literature of different countries. Motifs and

21  Reflections of Celtic Influence in Hildinavisen tales often have a wide distribution and it is often unsafe to suggest an intimate connection between various tales merely because each is a complex of similar motifs. A motif can be defined as the simplest form of a basic situation (Carney 1979: 48). Motifs or elements usually undergo adaptation. Roland Barthes states that narrative on the level of the story is translatable into different media and different cultural settings without fundamental damage. “It is the last layer, the discourse, which resists transference” (Barthes 1977: 121). I do not claim that the elements I am going to discuss below derive directly from certain Irish or Welsh stories, but I would like to emphasize that these elements are unique in Scandinavian balladry and it is not unlikely that they have sources other than Scandinavian. I discovered a number of parallels to the motifs in Shetlandic ballad in Celtic medieval literature. Speaking about the transference of literary motifs, or elements, it is necessary to touch upon another important issue, namely the change of literary medium, i. e. the difference of genre. Celtic sagas are prose with some poetic interpolations, whereas Scandinavian ballads are in verse. It is possible that stories were transmitted orally and the ballads were created from their motifs. Be that as it may, the sagas are not contemporaneous with the composition of the ballads. Prose texts usually provide space for countless details, while the poetic form of the ballad demands compactness and details tend to be obliterated. As Liestøl puts it, “instead of somewhat complicated content arrangement which is customary in fornaldarsögur, simpler and more popular fairytale motifs dominate in the ballads with the same subject”’ (Liestøl 1910: 272). Besides, in prose there is much less repetition and formula. Compared with the verse, dialogue in prose is natural and free in its movement, and is often conducted with swift-moving economy and terseness. Hurling of the Head As indicated, there is some evidence of strong links between Celtic sources and Scandinavian balladry in general. I would like to note some motifs that are especially prominent in Celtic literature and that also appear (certainly transformed) in this only ballad found in Shetland. The first one is the episode in Hildinavisen where, according to the summary of the ballad given by Low “the Earl is killed by Hiluge,

22  Rasa Baranauskienė who cut off his head and threw it at his lady, which, she says, vexed (!) her even more than his death, that he should add cruelty to revenge” (Low 1879: 113). The translation given by Low can be compared with the same lines from the ballad Hildinavisen: Hildinavisen 22 Nu fac an Iarlin dahuge Dar min de an engin gro An cast ans huge ei Fong ednar u vaxhedne mere mo. ‘Now the Earl got a deathblow – nobody could help / save him. He (Hilu­ge) cast his head into her (Hildina’s) lap (embrace, chest) and she was angered’ (Hægstad 1900: 6).

Two Irish sagas have episodes connected with beheading. One of them is Fled Bricrenn ‘Bricriu’s Feast’, the other is Scéla Muicce Meic Da Thó ‘The Story of Mac Dá Thó’s Pig’ which are reminiscent of the beheading episode in our ballad. Both these sagas have “the constellation of concepts that may be conveyed by headings such as ‘contention at the Celtic feast’, ‘the Celtic cult of the head’, and others of this type” (Koch 2000: 23–25). This feature was earlier described by Kenneth Hurlstone Jackson as the ‘head hunting and the beheading game’ ( Jackson 1964: 19–20, 35–37). The beheading motif can be called a stock motif and it is counted among the ones that belong to the earliest Celtic tradition. The earliest version of Fledd Bricrenn is found in the oldest Irish manuscript Leabhar na h-Uidre ‘The  Book of the Dun Cow’, written at Clonmacnoise about 1100 yet containing interpolations from 1250–1300. As it was proved in 1912 by the Irish palaeographer R. I. Best, the manuscript had been written by three different scribes.2 The story was first committed to writing in the 8th century – to judge by the language (O’Brien 1968: 68–69). Fledd Bricrenn contains the so called ‘beheading game episode’, where Cú Roi is beheaded three times, only

2 One of them was Maelmuire – murdered in 1106 by a marauder, probably a Viking.

23  Reflections of Celtic Influence in Hildinavisen to recover instantly. Concerning the sources of ‘the beheading game episode’ in the Icelandic Sveins rímur Múkssonar, Einar Ólafur Sveinsson rejects the idea that the motif in the Icelandic version could have been taken over from English or French sources (Sveinsson 1975: 134). The Irish story is closest to the Icelandic one, but the influence from Arthurian works indicates that the rímur could not have derived directly from the Irish tradition. A now lost source, possibly written in England, might therefore have served as an intermediary (Sveinsson 1975: 134). Orgain Mic Da Thó ‘Mac Dá Thó’s Slaughter’ is included in the list of prím-scéla (‘primary stories’) even before the period of our earliest manuscript text in the Book of Leinster. The list itself probably dates from the 10th century; but the tale is also mentioned in a poem by Flannacán Mac Cellaich who is said to have been slain by Norsemen in 896 (Chadwick 1968: 90). Scéla Muicce Meic Da Thó is also a very early story, probably composed in its present form about 800 AD. The setting of the story, and its link to Kildare, suggest that the author belonged to Leinster, and had inherited its fine heroic tradition. Scéla Muicce Meic Da Thó is highly sophisticated story which belongs to the early period of the Viking regime, and this may have done something to substitute laconic humour and a spirit of ripe burlesque for dignity and poetical beauty. The story is preserved in at least six manuscripts. The Book of Leinster, written about 1160 is the earliest. The text of the story is also found in Harley 5280, a manuscript written in the first half of the 16th century and now kept at the British Museum. These texts of the story are independent. They seem to be derived from a common source, which was a transcript of a previous version, believed from its language to date from about 800. Nora Chadwick assumed that Scéla Muicce Meic Da Thó also shows certain parallels with the Icelandic Bandamanna saga ‘The Story of the Banded Men’ (Chadwick 1957: 172), but these parallels are considered dubious by other scholars (Gísli Sigurðsson 1988: 93). Another episode from the same Irish story is paralleled in Njáls saga. Njáls saga has a description (Chapter 70) of an Irish dog Sámr, brought from Ireland as a gift by Ólafr Pá to Gunnar Hámundarson. The descriptions of legendary dogs in Irish stories are frequent, as for example the description of a dog brought from Spain in the opening lines of

24  Rasa Baranauskienė the above mentioned Scéla Muicce Meic Da Thó. If these two episodes in different family sagas are really connected with Scéla Muicce Meic Da Thó, then it seems that material of this Irish saga in one or other form was well known not just in Shetland, but also in Iceland. If Chadwick’s suggestion about these motifs is correct then it proves that the material of this particular Irish saga was well known in the area. Scéla Muicce Meic Da Thó refers to the  so-called curadmír  – ‘Hero’s Portion’. Diodorus Siculus, who already wrote about the Celts c. 60–30 BC, describing the behaviour of the Celts during feasts, mentions ‘the choicest portion’: “They honour the brave warriors with the choicest portion, just as Homer says that the chieftains honoured Ajax when he returned having defeated Hector in single combat. They also invite strangers to their feasts, inquiring of their identity and business only after the meal. During feasts it is their custom to be provoked by idle comments into heated disputes, followed by challenges and single combat to death” (Koch 1997: 11). The story Scéla Muicce Meic Da Thó tells about the rivalry of two heroes Cet Mac Mágach of Connacht and Conall Cernach of Ulster. Conall and Cet argue about the champion’s portion at the feast. At the end of the dialogue, Cet reluctantly acknowledges Conall to be a greater hero, regretfully adding that if a certain Anlúan were present, he would have challenged Conall: ‘He is present though’, cries Conall, who at this point takes the head of Anlúan which is hanging at his belt, and flings it at the opponent. ‘‘It is true’, Cet said, ‘you are even a better warrior than I. If Anlúan mac Mágach were in the house’, said Cet, ‘he would match you contest, and it is a shame that he is not in the house tonight.’ ‘But he is,’ said Conall, taking Anlúan’s head out of his belt and throwing it at Cet’s chest, so that a gush of blood broke over his lips’ (Koch 1997: 62).

Chadwick claims that “terse and humorous, with laconic brevity, it (the story) reminds us of the Icelandic sagas at their best. The dialogue in particular is masterly in its understatement and crisp repartee” (Chadwick 1968: 87).

25  Reflections of Celtic Influence in Hildinavisen The element with the head in the story was certainly capturing and probably used to make an indelible impression on the listeners of the story. The narrative aims at arousing and riveting attention and exciting interest, not at stimulating thought. The story-teller makes use of the element of surprise, of quick developments and dramatic moments. “He seeks to impress by rapid crescendo to a startling climax, and a shock, when Cet reluctantly gives precedence to Conall Cernach in the absence of Anlúan. There is more than a touch of humorous hyperbole in Conall’s throwing the head of Anlúan at Cet” (Chadwick 1968: 87–88). The story was evidently much liked in later times also, for it forms the subject of a number of independent poems. None of these seem to be based directly on the text of our saga. Chadwick suspects that the poems were inspired by a different version of the story (Chadwick 1968: 90). The motif is reused in Hildinavisen, but it is transformed and employed in a completely different context. The head of the dead husband is thrown at his wife. The motif occurs in a very dramatic and crucial moment of the story and perfectly serves its function to surprise and awaken interest and horror. This element is so particular that it is used twice – the second time at the very end of the ballad when Hiluge asks Hildina for mercy but she reminds him of having thrown the earl’s head at her and how much it had vexed her: Hildinavisen 34 Du tuchtada lide undocht yach Swo et sa ans bugin bleo Dogh casta ans huge I mit fung u vexmir mire mo ‘You thought I suffered not yet enough to see his body bleed, still you threw his head to my lap and I was vexed’ (Hægstad 1900: 9).

It is impossible to know how this motif penetrated into Hildinavisen, but it might have come through the Viking contacts with the aboriginal population in Shetland and Orkney, to whom the contents of the Irish

26  Rasa Baranauskienė sagas were known, because the story Scéla Muicce Meic Da Thó was coming into shape at the period of Viking invasions. It is clear that this motif in Hildinavisen is of Celtic origin, because in the Scandinavian balladry, except for Hildinavisen, it is used only in one more instance and is once more transformed and used in yet another different context. The severed head at the feast has a strong emotional effect (and the position of the episodes within the respective tales confirms that medieval authors felt the power of the device, much as we do now). Its currency may, therefore, be purely literary. Storytellers and writers knew an effective episode when they encountered it and simply reused the devise (Koch 2000: 35). As Marius Hægstad was the first to notice, we find fundamentally the same incident in the Faroese ballad Frúgvin Margareta (Hægstad 1900: 11). But in this ballad, it is not a full-grown man’s head, but a little child’s. Decapitation and throwing of the head are present just in Shetlandic Hildinavisen and Faroese Frúgvin Margareta (Hammersheib 1981: 93–120). The motif is Celtic, but the context is different. This element in the Faroese ballad might have been influenced by the Shetlandic ballad and reused in Frúgvin Margareta later – this time related to the killing of a child. The Faroese ballad is very likely to be much later, because it is clearly dominated by the Christian element. This motif was certainly very impressive and it does not appear in other Scandinavian ballads. As a result, an interesting amalgam of two cultures that merge together is achieved, since “the Irish Sea zone is hardly a culturally sterile environment” (Koch 2000: 27). The motif of the severed heads is well known in Iceland. When Gísli Sigurðsson deals with the nature of contacts on the Orkneys between Iceland and the Gaelic world after the age of settlement, he claims that the motif of the severed head comes into Icelandic from the Celtic world (Gísli Sigurðsson 1988: 12). There are numerous references in Icelandic sources about a head-cult of some sort and related folk beliefs. Many of these are believed to be due to the Gaelic influence and some are so well established that they are most likely to have developed within Iceland. Thus, in the Norse context, there is a group of motifs, atypical as they are of the Norse tradition and familiar from Celtic sources. Decapitation, the preservation of the severed head, its association with

27  Reflections of Celtic Influence in Hildinavisen a well, its powers of prophecy as well as otherworldly knowledge are all features which recur in the Celtic tradition and belief. All the evidence suggests that this episode in Norse mythology, if not a direct borrowing from a Celtic source, at least owes its presence in the Norse tradition to detailed knowledge on the part of the story-teller of such beliefs amongst the Celts (Ross 1962: 41). Severed talking heads at feasts appear in many Irish stories, particularly in the Finn Cycle. Bruiden Átha Í ‘The Quarrel at the Ford of the Yew Tree’ (Meyer 1893: 24), Aided Find ‘The Death of Finn mac Cumaill’ (Meyer 1897: 464–5) and Sanas Cormaic ‘Cormac’s glossary’ (the glossary of Bishop Cormac mac Cuillenáin, year 908) (Meyer 1912: xix–xx) contain episodes where a severed head demands its share of food. Severed talking heads in Old Icelandic material are to be found in Eyrbyggja saga, ch. 43 and Þorsteins þáttr b jarmagns, ch. 9. Severed heads of enemies appear in Grettis saga, ch. 82., Bjarnar saga Hítd lakappa, ch. 32, Fóstbr ðra saga, ch. 18, and Ljósvetinga saga, (Þórarins ­þáttr). Supernatural qualities are also attached to heads in Ólafs saga Tryggvasonar, ch. 28/19, Eyrbyggja saga, ch. 27, and Njáls saga, ch. 157 (i. e. the head of King Brjánn) (Gísli Sigurðsson 1988: 81). It is important to point out that a similar tradition of severed heads is also found in Orkney, but here it is slightly different, though there are certain parallels even with Scéla Muicce Meic Da Thó. In Orkneyinga saga (chapter 5), Sigurðr, the first Earl of Orkney, defeats the Scottish Earl Melbrikta nicknamed tönn ‘tooth’ in a battle, cuts his and his followers’ heads off, attaches them to the saddle and gallops triumphantly away. Unfortunately for Sigurðr, Melbrikta’s tooth, sticking out of the severed head’s mouth, wounds Sigurðr’s calf, causing a deadly infection. Note that both the Shetlandic ballad and Orkneyinga saga involve the Earl of Orkney. Besides, Melbrikta is an Irish name, meaning ‘devotee to St. Brigit’. The very custom of using heads as a token of triumph and even hanging them on horses was common enough among the Celts, examples of which can be found in numerous sources (Chadwick 1970: 49–50; Coch and Carey 1997: 12). One of the most famous Irish sagas Táin Bó Cúailnge ‘The Cattle-Raid of Cooley’ contains numerous episodes about Cú Chulainn galloping away with a bunch of heads tied to his horse.

28  Rasa Baranauskienė The distinctive element in Orkneyinga saga, however, is that the headepisode is connected with revenge. An Old Irish parallel to this combination of motifs can be found in Aided Chonchobuir ‘The Death of Conchobhar’ ( Jackson 1971: 53–56), dated in the 9th century. Aided Chonchobuir also describes a feast with disputes and contentions among the Ulstermen. In this story, a ball made out of the Leinster King Mesgegra’s brain and used by the Ulstermen to boast about the victory, is stolen by a Connachtman, Cet: “He snatched the brain from the hand of one of them [buffoons] and carried it off with him, for Cet knew that it was foretold that Meis-Geghra would avenge himself after his death” ( Jackson 1971: 54). Eventually, this particular brain is thrown at the Ulster King, Conchobhar Mac Nessa: “Cet fitted Meis-Geghra’s brain into the sling, and slung it so that it struck Conchobhar on the top of his skull, so that two-thirds of it were in his head, and he fell headlong on the ground” ( Jackson 1971: 54). The ball enters his head but does not cause his death until several years later, when Conchobhar receives the news of Christ’s crucifixion. Then the ball falls out of his head, leaving a hole for the blood to gush forth, whereupon Conchobhar dies, is baptized in his own blood and becomes the first Irishman to go straight to Heaven. The pattern of revenge is quite complicated in the Irish story and not as straightforward as in Orkneyinga saga where the full-sized head kills the actual killer, soon after having been separated from the body (Gísli Sigurðsson 1988: 45–46). The similarities nevertheless lead Bo Almqvist to conclude: “One need not assume that the tale about Mesgegra’s brain is the direct source of the Melbrikta episode in Orkneyinga saga, but some such Gaelic story, perhaps in a more primitive form and without hagiographic ingredients, seems likely to lie behind it” (Almqvist 1981: 99). In the above mentioned Irish saga Aided Chonchobuir we also have a hurling episode, but it is not a whole head that is thrown, but a ball made out of the brain. However the similarity of the motifs and the motivation of this action, i. e. revenge, are obvious. It is also worth mentioning that the very same Cet plays a crucial role in causing the King Conchobhar’s death, i. e. he is the thrower. In Scéla Muicce Meic Da Thó, the head is thrown at him. It is also very important that Cet’s rival, Conall, mentioned in Scéla Muicce Meic Da Thó acts as one of Cet’s rivals also in

29  Reflections of Celtic Influence in Hildinavisen Aided Chonchobuir. It might happen that some Old Irish stories about the rivalry between the Ulstermen and Connachtmen and Cet’s destiny, where severed heads (alternatively – balls made of brain) appear and are being hurled forth in order to cause the rival’s death or in revenge, were well known in Orkney and Shetland and thus were paralleled in various Scandinavian texts. “King and Goddess” Theme in Hildinavisen The narrative of Hildinavisen is particular and divergent compared to other heroic Scandinavian ballads. The story evolves around a woman, who takes revenge for her husband in a very particular way. This Shetlandic ballad, or rather its framework, is reminiscent of an adaptation of the Celtic “King and goddess” theme. The Celtic world shared with many other ancient cultures the mythic model of the royal rule – hieros gamos, or sacred marriage. According to this model, successful and prosperous government of society was the outcome of the union between female and male elements, between the goddess of the land and its sovereign (Herbert 1992: 264). In the universe of early Irish mythology, the female deity was the embodiment both of the physical land and of its dominion (Herbert 1992: 56). The feminisation of the land is amply in evidence in the sovereignty myth (Herbert 1992: 57). In its Celtic setting, the myth is represented primarily in sources from both Gaul and Ireland. The abundance of the stories containing the “King and goddess” theme in both Irish and Welsh medieval literature is a wellknown fact and sacred marriage imagery has been a constantly recurring theme in Irish literature throughout the ages (Breatnach 1953: 321–36). The durability of the theme as a literary topos is surprising. Every time the motif appears, the story is different, but we still can notice an on-going continuity which inevitably implies the transformation of each and every story. The Gaulish epigraphic and iconographic evidence belongs to the period between ca. 500 BC and 400 AD. Written sources referring to the Gaulish society were produced by Greek and Roman observers (Mac Cana 1970: 16–17). In Ireland literary evidence belongs to the period from about the 7th century AD onward. In the case of early

30  Rasa Baranauskienė Irish narrative, the hypothesis is that we are dealing with mythology refracted through literature (Ó Cathasaigh 1993: 128). What is remarkable, however, is the persistence and vigour of these concepts in the tradition of the only Celtic society which remained relatively untouched by the Roman civilization (Mac Cana 1970: 121). The iconographic imagery of a foreign consort of the goddess of the land finds a literary reflex in the story of the foundation of Massilia (Marseilles). It relates that the Gaulish king’s daughter, in the act of proffering a symbolic marriage libation to her intended spouse, bestows the drink on the newly-arrived foreigner (Herbert 1971: 265). We find very similar stories in early Ireland where a goddess validates the ruler through the act of marriage. In medieval Irish literature, we have narratives relating to two Medbs – Medb of Cruachu and Medb Lethderg of Leinster – both of them select and validate their royal spouses through marriage. The theme persisted in Celtic territories almost unchanged in its lineament and in its influence. The concept of a female bestowing the right to rule on male sovereigns remained as a rather stable and yet shifting phenomenon, so the preservation of the myth in the literary sources of the early Christian period seem to trace the survival of its narrative realization. The most famous text Baile in Scáil, ‘Phantom’s frenzy’, dated from the early 11th century AD (Gerard Murphy holds that there is an earlier stratum in the text, possibly from the ninth century [Murphy 1937: 143]) portrays a vision of the pan-Celtic god-king Lug enthroned in iconic fashion beside his female consort. She is instructed to bestow the drink of sovereignty on a succession of rulers destined to be kings. The imagery is strongly reminiscent of the Gaulish representations (Herbert 1971: 267). Here the significance of the drink is prominent. There are many other Irish equivalents where the goddess destroys the unrighteous and confirms the right one as a king. The underlying pattern of the stories is the same. There are special elements constitutive of the account, i. e. brothers claiming kingship, a hunt in the wilderness, a disguised queen and an apparently repugnant sexual union, which have been noted by J. de Vries. According to J. Carey, there are some recurring elements

31  Reflections of Celtic Influence in Hildinavisen like hunt and / or wandering, woman dispensing a drink, woman who appears in different forms. These elements (not necessarily all of them) recur in the famous legends told of Niall Noígiallach and Lugaid Laígde (De Vries 1961: 120). Another similar story is the legend of Macha Mongruad. In the legends of Niall and Lugaid, the true claimant is united with the goddess and in the case of Mongruad she subjugates the unworthy (Carey 1983: 69). Mongruad is an example of the terrible aspect of the Sovereignty goddess (Carey 1983: 263–75). Not all encounters with the Sovereignty goddess are equally benevolent. A notable feature of the system was the dual aspect of the goddess. The figure of the sovereignty could appear repulsive or beautiful. “Death and slaughter were the reverse sides of the personifications of growth and fertility” (Carey 1983: 268). There is a wide range of Gaulish and Irish narratives, where instead of being confirmed as a king by the female divinity, the unsuitable ruler is destroyed. In some occasions Sovereignty goddess displays her twofold character: sinister and aggressive on the one hand, beautiful and prosperous on the other (Carey 1983: 268). The  tale which is of great importance to the  investigation of Hildinavisen is one of the earliest Celtic tales containing the “King and goddess” theme with a negative outcome where the goddess acts as terrifying and malevolent. The setting of the story is similar to that of the above-mentioned foundation story of Masillia. But almost the same scenario can lead to two different consequences. One version of the story comes from Asia Minor (around modern Ankara, Turkey). The story is registered by Mestrius Plutarchus (Plutarch c. 46–127) in section XX of his Moralia in the chapter De Mulierum Virtutibus, ‘On the Bravery of Women’, 257–8, called “The Poisoned libation: the Love Triangle of Sinatus, Sinorīx, and the High Priestess Camma”. This work of Plutarch appears in pp. 471–581 of Vol. III of the Loeb Classical Library’s edition of the Moralia, first published in 1931. Polyaenus (the middle of the 2nd century AD) drew freely from Plutarch’s Moralia to embellish his Strategemata. Sinorīx means ‘old king’, Camma probably means ‘evil woman’ and Sinātus means ‘the one with good ancestry’. Galatia here means the Celtic domain founded in Hellenistic times in central Asia Minor. Features of the narrative – including a queen closely connected

32  Rasa Baranauskienė with a goddess, a honey drink that proves poisonous, an unnatural death instead of a wedding feast, a chieftain set in a chariot as his relatives prepare his tomb, a love triangle terminating in a fateful chariot ride and kinslaying as the prelude to the downfall of the king, a woman who brings great evil to those close to her through no fault of her own – resonate widely through the Celtic literary traditions and may be viewed as elements in its inherited preliterary substance (Koch 1997: 34). In order to compare the story about Camma and Hildina I give here the Plutarch’s story in full. Translation is made by John Carey in The Celtic Heroic Age. XX. Camma Sinātus and Sinorīx, distant kinsmen, were the most powerful of the tetrarchs of Galatia. Sinātus had a young wife named Cammā, much admired for her youth and beauty, but still more remarkable for her virtues. For she was not only modest and affectionate, but also shrewd and courageous, and fervently beloved by her servants on account of her compassion and her kindness. She was further distinguished by her office as priestess of Artemis, the goddess whom the Galatae most revere, and was always to be seen at the solemn processions and sacrifices, magnificently attired. Sinorīx fell in love with her. Unable to possess her either by persuasion or by force while her husband lived, he did a dreadful deed: he killed Sinātus treacherously. Not long thereafter he proposed to Cammā, who was now living in the temple. She was biding her time, and bore Sinorīx’s crime not with pathetic weakness but with a keen and foreseeing spirit. He was importunate in his entreaties, and proffered arguments not entirely implausible: he claimed that he was a better man than Sinātus and had killed him for no reason except his love for Cammā. Even at first, her refusals were not too harsh, and in a little while she seemed to soften. (Her relatives and friends were also pressuring and seeking to force her to accept him, hoping themselves for the favour of the mighty Sinorīx.)

33  Reflections of Celtic Influence in Hildinavisen At last she yielded, and sent for him so that the compact and the vows might be made in the presence of the goddess. When he arrived she received him affectionately. She led him to the altar, poured a libation from a drinking-bowl, drank some herself, and told him to drink the rest. It was a drink of milk and honey [melikraton], with poison in it. When she saw that he had drunk, she cried aloud and fell down before the goddess. ‘I bear witness to you, most glorious spirit,’ she said, ‘that it is for the sake of this day that I have lived since Sinātus’s murder, in all that time taking pleasure in none of the good things of life, but only in the hope of justice. Having attained this, I go down to my husband. As for you, most impious of men, your relatives can prepare your tomb, instead of your wedding and bridal chamber.’ When the Galatians heard this, and felt the poison at work in him and penetrating his body, he mounted his chariot as if the tossing and shaking might do him good; but forthwith he desisted, got into a litter, and died in the evening. Cammā survived through the night: learning of his death, she passed away cheerfully and gladly.

The scenario is almost the same as in the Shetlandic ballad. However, it is clear that Celtic motifs reused in the ballad have nothing to do with their mythological aspect, it is not rationalisation of the myth. In case of Hildinavisen, it is just a borrowing of the narrative. In the story Camma is depicted as a mortal female, though she is connected with the goddess Artemis, goddess of hunting, (‘whom the Galatae most revere’) being her priestess. In Hildinavisen and in the story of Camma we have a typical Celtic love triangle, where the heroine’s husband is killed by a jealous rival who discloses his intentions to marry the widow straightaway. Compare to Hildinavisen: Hildinavisen 23 Di lava mir gugna Yift bal yagh fur o landi Gipt mir nu fruan Hildina Vath godle u fasta bande.

34  Rasa Baranauskienė ‘You let me get married if she will follow me from the country, give me now lady Hildina with gold and betrothal’ (Hægstad 1900: 6).

With regard to marriage, women in both stories seem to act on their own free will. In Hildinavisen, Hildina’s father asks Hilugi to wait until the child is a bit older and then leaves to Hildina the right to decide. In the story of Camma though Camma seems to be urged to marry Sinorīx by her relatives, she can finally decide herself. Hiluge as Sinorīx is clearly a wrongful king, not destined for kingship. The scenario of the story is the same as in the stories with the “King and Goddess” theme that have a positive outcome, but the drink turns out to be poisoned. The symbolism of the sacred drink is absolutely transparent in the story of Camma. An emphasis on the drink is also very lucid in the Shetlandic ballad. Hildina concedes to marry Hiluge, but asks to be allowed to serve the wine. Her father allows her on condition that she will not think about the Earl. Hildina answers that even if she thought about the Earl, she would not serve any harmful drink to her father: Hildinavisen 26, 27, 28 Nu Hildina on askar feyrin Sien di gava mier live Ou skinka vin Ou guida vin. Duska skinka vin, u guida vin Tinka dogh eke wo Iarlin an gougha here din. Watha skilde tinka Wo Iarlin gouga herè min Hien mindi yagh inga forlskona Bera fare kera fyrin min. ‘Now Hildina asks her father – Give me permission to dispense the wine, to pour the wine.

35  Reflections of Celtic Influence in Hildinavisen You shall dispense the wine and pour the wine, though do not think about the earl, your good lord. Though I will think about the earl, my good lord, for that I would not serve any harmful drink to my dear father’ (Hægstad 1900: 7).

Hægstad takes the word forlskona as a compound in genitive case – *fárskonnu, composed of the word fár, meaning ‘harm’ and kanna, ‘vessel’. Later on, Hildina serves a drink to her father and everybody else. In this case, the drink is called mien. It corresponds to the Old Norse word mjoðr ‘a drink made out of milk and honey’. In the Celtic story, the drink is called melicatron and is made of milk and honey. It is clear from the text that Hildina ‘infuses a drug’, but it is not clear what kind of drug it is: Hildinavisen 25 Hildina liger wo chaldona U o dukrar u grothè Min du buga till bridleusin Bonlothir u duka dogha. ‘Hildina lies in the tent, her eyes are dark of crying, and before she is called to the wedding ceremony, she infuses poison into the drink’ (Hægstad 1900: 7).

Here the ending of the story is different – in the Celtic tradition, the man dies, but in this story the main hero is burned alive as in so many different Icelandic sagas. In general, the serving of ale and mead in Scandinavian ballads is traditional, but here the tradition is modified with poisoned libation and combined with a typical Celtic triangle. The burning, on the contrary, seems to be in itself a particularly Scandinavian motif, which often appears in various Icelandic sagas. In the end, when Hiluge asks Hildina to pity him, she again reminds him of the throwing of the earl’s head at her. Again, this seemingly Celtic motif is exploited here with a new strength. The motif of a king receiving a drink from a beautiful woman in Scandinavian literature is not unique to Hildinavisen. Generally stories

36  Rasa Baranauskienė with this motif are held to be closely connected with the Irish tradition like for example stories about the Norwegian King Haraldr hárfagri, found in Hálfdanar saga svarta (chapter 8) (Aðalbjarnarson 1941: 84–93) and Flateyjarbók (Flateyjarbók, 1860–1862, 564–66). The King as a young man follows Finn or Dofri into a supernatural fosterage where he receives a cup of mead from his fosterer’s beautiful daughter and is promised a sovereignty on his departure. Chadwick also maintains that the appearance of a Finn, a famous Irish hero, in the Icelandic version is significant (Chadwick 1957: 192) and demonstrates Celtic and Scandinavian contacts. Close relationship of Harald’s supernatural experiences to those of the Irish High-King Conn Cétchathach, and still more to those of Conn’s descendant, Niall Noígiallach, as well as their relations with the maiden calling herself the flaithiusa h-Erenn ‘the Sovereignty of Ireland’ are already well established (Chadwick 1957: 192). Conclusions The presence of Celtic elements in the only surviving ballad in Norn language Hildinavisen, which have been created in continually changing linguistic, social and cultural conditions, indicates a certain degree of contact with the Celtic population. The marine societies of Shetland and Orkney Islands generated various stories, where motifs and elements travelled in various directions. Hildinavisen is certainly of Western Scandinavian origin, but it contains or rather is adorned with Celtic motifs. The meeting of these two traditions – Celtic and Scandinavian – gave a peculiar and productive treatment of certain narratives. The analysis of the text of Hildinavisen indicates that on the level of the story Hildinavisen borrows substantially from the Celtic narrative tradition. Various motifs are externally imposed and later on adapted in Scandinavian material. Especially prominent is the appeal of the King and Goddess theme. But if it is a borrowing, it is by no means direct, because Celtic motifs and elements are adapted, transformed and melted in the text of the Scandinavian ballad. The presence of some Celtic motifs, such as throwing of the head, which is one of the favourite devices in the Irish sagas, or the King and goddess theme have wide ramifications in Celtic literature and were probably particularly prominent in the oral tradition.

37  Reflections of Celtic Influence in Hildinavisen It is possible that there was some kind of Orkney version of the story later on transmitted to Shetland, since the main hero is the Earl of Orkney. This story might have served as an intermediary between Hildinavisen and the Celtic tradition, since Orkney have been one of the important channels for the transmission of the Celtic elements and Shetland must have been on the route of these contacts. Bearing in mind the historical modes of habitation, as the sea was worn deep with boats, the favourite pastimes and working activities were connected with sailing and rowing, usually accompanied by ballads. Thus the ballad was preserved in the bound language for a long time. This rowing and singing together catalysed the appearance of new motifs, which were designated to address issues specific to the unique conditions of this region. Appendix I. Contents of Hildinavisen after Low Low wrote about the content of the ballad (Anders. page 113): “A literal translation of the above (the ballad) I could not procure, but the substance is this:” An Earl of Orkney, in some of his rambles on the coast of Norway,3 saw and fell in love with the King’s daughter of the country. As their passion happened to be reciprocal, he carried her off in her father’s absence, who was engaged in war with some of his distant neighbours (v. 1–3). On his return, he followed the fugitives to Orkney, accompanied by his army, to revenge on the Earl the rape of his daughter (v. 7). On his arrival there, Hildina (which was her name), first (!) spied him, and advised her now husband to go an attempt to pacify the King (v. 9). He did so, and by his appearance and promises brought the King so over as to be satisfied with the match (v. 12). This, however, was of no long standing, for as soon as the Earl’s back was turned (!) a courtier, called Hiluge, took great pains to change the King’s mind, for it seems Hiluge had formerly hoped to succeed with the daughter himself (v. 15–16). His project took, and the matter came to

3 The mark for something that Low says, but which is actually not found in the ballad.

38  Rasa Baranauskienė blows (v. 16–18); the Earl is killed by Hiluge, who cut off his head and threw it at his lady, which, she says, vexed (!) her even more than his death, that he should add cruelty to revenge (v. 22). Upon the Earl’s death, Hildina is forced to follow her father to Norway, and in a little time Hiluge makes his demand to have her in marriage of her father; he consents, and takes every method to persuade Hildina, who with great reluctance, agrees upon condition that she is allowed to fill the wine at her wedding (v. 26). This is easily permitted (v. 27), and Hildina infuses a drug (v. 25) which soon throws the company into a dead sleep, and after ordering her father to be removed, set the house on fire (v. 29–30). The flame soon rouses Hiluge, who piteously cries for mercy, but the taunts he had bestowed at the death of the Earl of Orkney are now bitterly returned, and he is left to perish in the flames (v. 31–34).

Bibliography Almqvist, Bo. 1981. “Scandinavian and Celtic Folklore Contacts in the Earldom of Orkney.” Saga-Book, vol. XX (1978–81). Viking Society for Northern Research, 80–105. Barnes, Michael P. 1996. “Jakob Jakobsen and the Norn Language of Shetland.” Shetland’s Northern Links. Language & History. Edited by Doreen J. Waugh. Scottish Society for Northern Studies, 1–15. ———. 2000. “Norse and Norn.” Languages in Britain & Ireland. Edited by Glanville Price. Oxford: Blackwell Publishers. Barthes, Roland. 1977. “Introduction to the Structural Analysis of Narrative.” Image, Music, Text. Glasgow: Fontana / Collins. Bjarni Aðalbjarnason. 1941. Heimskringla I. Íslenzk fornrit, 26, 84–93. Reykjavík: (Hið íslenzka fornritafélag). Breatnach, Risteard A. 1953. “The Lady and the King: A Theme of Irish Literature.” Studies, 42. Dublin: Dublin Institute for Advanced Studies, 321–36. Carey, John. 1983. “Notes on the Irish War-Goddess.” Éigse, 19. Dublin: National University of Ireland, 263–75. Carney, James. 1979. Studies in Irish Literature and History. Dublin: Dublin Institute for Advanced Studies. Chadwick, Nora K. 1953–1957. “Literary Tradition in the Old Norse and Celtic World.” Saga Book, vol. XIV, part II. Viking Society for Northern Research, 164–99. ———. 1968. “Scéla Muicce Meicc Da Thó.” Irish Sagas. Edited by Myles Dillon, Dublin and Cork, the Thomas Davis Lecture Series, published in collaboration with Radio Telefís Éireann by Mercier Press.

39  Reflections of Celtic Influence in Hildinavisen ———. 1970. The Celts. Penguin Books. Chesnutt, Michael. 1968. “An unsolved problem in Old Norse-Icelandic Literary History.” Mediaeval Scandinavia, 1. Odense: Odense University Press, 122–37. Davíð Erlingsson. 1975. “Illuga saga og Illuga dans.” Gripla. Ritstjóri Jónas Krist­ jánsson. Reykjavík: Stofnun Árna Magnússonar, 9–41. De Vries, Jan. 1961. Keltische Religion. Stuttgart. Einar Ólafur Sveinsson. 1959. “Celtic Elements in Icelandic Tradition.” Béaloideas, 25. Dublin, 3–24. ———. Ólafur Sveinsson. 1975. “Löng er för.” Studia Islandica, 34. Reykjavík. Flateyjarbók I–II. 1860–1862. Christiania. Flom, George Tobias. 1925. “Breaking in Old Norse and Old English: With Special Reference to the Relations between Them.” Language, 13. Washington: Lin­guistic Society of America. Gísli Sigurðsson. 1988. Gaelic Influence in Iceland. Historical and Literary Contacts. A Survey of Research. Reykjavík: Bókaútgáfa menningarsjóðs. Grundtvig, Sven. 1853. Danmarks Gamle Folkeviser. Kjöbenhavn: Thieles Bog­ trykker. Herbert, Máire. 1971. “Transmutations of an Irish Goddess.” The  Concept of the Goddess. Edited by Sandra Billington and Miranda Green. London and New York: Routledge, 141–150. ———. 1992. “The  Universe of Male and Female: A Reading of the  Deirdre Story.” Proceed­ings of North American Congress of Celtic Studies. Halifax: Nova Scotia, 53–64. Hammerheib, Venceslaus Ulricus. 1981. F røisk anthologi; Text samt Historisk og gramatisk indledning med understøttelse af Carlsbergfondet. København: S. L. Møllers. Hægstad, Marius. 1900. “Hildinakvadet med utgreiding um det norske maal paa Shetland i elder tid.” Videnskabsselskabets Skrifter. II Historisk-filosofiske Klasse. No. 2. Christiania, I Kommission hos Jacob Dybwad: Grøndahl & Søns Bogtrykkeri. Holtsmark, Anne. 1939. “Vefr Darraðar.” Maal og Minne, 74–96. ———. 1964. “Kongespeillitteratur.” Kulturhistoriskt Lexikon för nordisk Medeltid II. Malmö. Jackson, Kenneth Hurlstone. 1964. The Oldest Irish Tradition: A Window on the Iron Age. Cambridge: Cambridge University Press. ———. 1971. A Celtic Miscellany. Penguin Books. Koch, John T. and Carey, John. 1997. The Celtic Heroic Age. Literary Sources for Ancient Celtic Europe and Early Ireland and Wales. Andover, Massachusetts: Celtic Studies Publications. Koch, John T. 2000. “Fled Bricrenn’s Significance within the Broader Celtic Context.” Fled Bricrenn: Reassessments. Edited by Pádraig Ó Riain. Dublin: The Irish Texts Society (Subsidiary Series No. 10).

40  Rasa Baranauskienė Landstad, Magnus Brostrup. 1854. Norske Folkevisor. Christiania. Liestøl, Knut. 1910. Syn og Segn. Oslo. Liestøl, Knut og Moe, Moltke. 1958. “Folkeviser I.” Norsk Folkediktning VI. Ny utgave ved Olav Bø og Svale Solheim. Oslo. Low, George. 1879. A Tour through the Islands of Orkney and Shetland 1774. Kirkwall: William peace & Son. Mac Cana, Proinsias. 1970. Celtic Mythology. London and New York: Hamlyn. Meyer, Kuno. 1893. “Two tales about Finn (Bruiden Atha and Tucait fhagbála in fessa) from Stowe MS 992.” Revue Celtique, vol. 14, 241–249. ———. 1897. “The Death of Finn mac Cumaill.” Zeitschrift für Celtische Philologie, 1, 462–465. ———. 1912. “Sanas Cormaic from the Yellow Book of Lecan.” Anecdota, 4. Murphy, Gerard. 1937–1956. Early Irish Lyrics. Oxford: Oxford University Press. O’ Brien, Maureen. A. 1968. “Fled Bricrenn”. Irish Sagas. Edited by Myles Dilbon. The Thomas Davis Lecture Series. Dublin and Cork: published in collaboration with Radio Telefís Éireann by the Mercier Press. Ó Cathasaigh, Tomás. 1993. “Mythology in Táin Bó Cúailnge.” Studien zur Táin Bó Cúailnge. Heraus­gegeben von Hildegard L. C. Tristram. Tübingen: Gunter Narr Verlag. Rendboe, Laurits. 1987. “Det gamle shetlandske sprog; George Low’s ordliste fra 1774.” North-western European Language Evolution, Supplement vol. 3. Odense: Odense universitetsforlag. Ross, Anne. 1962. “Severed Heads in Wells: an Aspect of the Well Cult.” Scottish Studies, vol. VI, 31–48. Sibbald, Sir Robert. 1771. The Description of the Isles of Orkney and Zetland. Edinburgh: Stevenson. Wallace, James. 1700. An Account of the Islands of Orkney. London: Jacob Tonson.

On Translating Icelandic Sagas into Modern Norwegian – the Case of Brennu Njáls Saga Jan Ragnar Hagland Norwegian University of Science and Technology

Gathered as we are here in Vilnius, on the splendid occasion of the 200th anniversary for the publication of professor Joachim Lelewel’s Edda Skandinawska, it might be pertinent to look at parallel instances of mediating Old Icelandic literature by way of translation elsewhere. It is, as we all know, a fact that the speaker of any language – modern Icelandic excepted – who is not specializing in Old Norse, has to resort to translations in order to appreciate the literature handed down to us in that particular language. As this literature contains some of the masterpieces of medieval European literature, translations are, of course, important. This is so, needless to say, also when translations into a quite closely related language such as modern Norwegian are concerned. I shall, then, try to look at some aspects of Norwegian saga translations. In order not to exceed all reasonable time limits, I shall confine the present exposition and discussion to the translation history of one of the major works in Old Icelandic literature – the Brennu Njáls Saga. The translation history of a text such as this, is of course, closely related to its historical reception, as is, no doubt, the edition we are celebrating in the present seminar. The Njáls Saga has a fairly long translation history in Norway – a history which each individual translated version unveils, I think it is fair to say, aspects of the historical reception of the saga, as well as aspects of the state of the art where saga research in general is concerned. We will return to the former of these two points – the latter shall be left at that here, even if this is something which may be argued. The first to make a point of having made a translation into Norwegian rather than into Danish – the common written language used in Denmark and Norway at the time – was Karl L[inné]. Sommerfelt, who made a translation into Dano-Norwegian, which he published as Approaching the Viking Age. Proceedings of the International Conference on Old Norse Literature, Mythology, Culture, Social Life and Language. Edited by Ērika Sausverde and Ieva Steponavičiūtė. (Scandinavistica Vilnensis 2). Vilnius University Press, 2009. https://doi.org/10.15388/ScandinavisticaVilnensis.2009.2.3 Copyright © 2019 Authors. Published by Vilnius University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

42  Jan Ragnar Hagland an appendix to a periodical for the enlightenment of the people called Folkevennen in 1871. As a translation into Danish had been re-published only nine years earlier (Petersen 1862), the reasons given by Sommerfelt for publishing a new version are interesting both ideologically and linguistically. Sommerfelt seems compelled to excuse himself for publishing a new translation so soon after Petersen – he had done so, he says, because professor Petersen’s translation, in spite of its indisputable qualities, could not really satisfy the tastes of a Norwegian reading public. This was no fault of the translator – it had to do with the Danish language. Danish and Norwegian, Sommerfelt says, are, even if close, two different languages representing two different nationalities. ‘In consequence’, he goes on to say, ‘I do not think that a Norwegian reader in general will be satisfied if he has to make the acquaintance of this particular literature through the medium of the Danish language. Even if he does not know the original language, the reader will easily be struck by the impression that the Danish language is too weak to mediate the force and vigour, the exuberance of expression, the magnificence and boldness of characters and scenes typical for the saga’.1 A modern reader would most certainly experience difficulties in finding support in the translated text for this programmatic statement of linguistic difference between Dano-Norwegian and Danish, even if differences may, of course, be observed on closer investigation. What is more important in this context is, in my opinion, Sommerfelt’s obvious intention to associate the sagas – the literature of the Golden Medieval Age – with the Norwegian language rather than Danish. This is indeed an expression of a growing awareness of a specific Norwegian identity,

1 “Professor Petersens Oversættelse vil, hvilke Fortjenester den end uimodsigelig besidder, neppe ret kunne tilfredsstille en norsk Læser. Dette er ikke Oversætterens Feil, men det danske Sprogs. Dansk og Norsk vise sig, hvor nær de end staa hinanden, dog i visse Henseender at være to forskjellige Sprog, ligesom de representerer forskjellige Nationaliteter […] Jeg tror derfor, at en norsk Læser i Almindelighed ikke vil føle sig tilfredsstillet ved at gjøre Bekjendtskab med denne Litteratur gjennem det danske Sprog. Selv om han ikke kjender Originalen, vil det let paatrænge sig ham, at det danske Sprog er for vegt til at gjengive den Kraft, Fynd og Kjernefylde i Udtrykket, den Storslagenhed og Djærvhed i Karakterer og Scener, som er Sagaen egen…” (Sommerfelt 1871, III–V.).

43  On Translating Icelandic Sagas into Modern Norwegian nourished no doubt by the ideas of romanticism – an ideological basis also, and perhaps even more so, for the activity of translating sagas into the alternative written language in Norway – the then newly coined “Landsmaal”. We will return to this later on. It is suffice here just to point out the fact that the first saga to be translated into Ivar Aasen’s “Landsmaal” (by himself) – the Fridtjovs Saga – was published on the initiative of Folkevennen, the very same periodical that published Sommerfelt’s translation of the Njáls saga. After Sommerfelt’s translation, a quarter of a century passed before the next translation of Njáls Saga appeared in Norway. As vaguely indicated above, a project of translating sagas into Ivar Aasen’s then recently established “Landsmaal”, can be observed through the latter half of the 19th century. Ideologically this was, it appears, part of a larger process of language planning in which the translating of sagas constituted, it seems fair to say, a retrospective dimension in the making of a new literary language (cf. Hagland 2003, 44–45). After Ivar Aasen’s translation of Fridtjovs Saga in 1858, some others had tried to forge translated saga texts into the new written idiom. The translations of saga texts into this new written standard should be looked upon, in my opinion, as a conscious search for linguistic and literary models, or patterns perhaps, in the process of making it a literary language. As such, the translations of sagas – and biblical texts for that matter – represented an archaizing element on which we shall not elaborate in the present context (cf. though Hagland 2003, 45–47). Anyway – in this process of making a literary language, a certain number of attempts at translating sagas had already been undertaken when Olav Torsson Aasmundsstad in the mid 1890s ventured the difficult task of translating Njáls saga into the new “Landsmaal”: Njaala eller Soga um Njaal Torgeirsson og sønerne hans published in 1896–97. Since then a parallel course of translations and re-editions of this particular saga has been a distinct feature in the history of literary translations into Norwegian – one into Dano-Norwegian, later “Riksmål” and “Bokmål” the other into “Landsmaal”, later “Nynorsk”. It should be added here that Jón Karl Helgason in his interesting study The Rewriting of Njáls saga seems to overlook the importance of the translations into “Landsmaal” in his discussion of Njáls Saga and

44  Jan Ragnar Hagland Norwegian liberation – language and nationality ( Jón Karl Helgason 1999, 101–116). Before proceeding any further we shall briefly recapitulate the main stages in the subsequent history of Njál’s Saga translations into the two Norwegian written standards: Sommerfelt’s translation into DanoNorwegian was replaced – if we may use an expression like that in this context – by Fredrik Paasche’s translation into “Riksmål” in 1922 (cf. bibliography), an influential version of the text which was reissued in 1986 and again in 19992. The next step along this line was Hallvard Lie’s translation of 1941. This is the most complete version of the existing translations into modern Norwegian – only very minor omissions can be observed in the genealogies. This translation was, however, reedited and reworked into a much abbreviated version (see below) in 1951, published in the popular series of “Hjemmenes boksamling”. A somewhat more comprehensive version was published in 1954, in the series “Islandske ættesagaer” edited by Hallvard Lie. The former of the two is probably the most widely distributed version of Njáls Saga among the reading public in Norway ever – (exact numbers of printed copies are, however, not available). Aasmundstad’s translation appears to have been out of print and seems, for linguistic reasons, to have been somewhat out of date by the mid 1920s. To cover up for this Knut Liestøl, the famous saga scholar, published a linguistically revised version in 1928. Even if this revised version was based upon an edition of the saga in the source language more updated than the one Aasmundsstad had at his disposal, no major changes in the text as a whole were made.3 A third revision – or retranslation rather – was made by Knut Liestøl’s son Aslak in 1961, a version that – as far as the extent of the text is concerned – deviates somewhat from the 1928 version. It was reprinted in 1975, in the series “Norrøne bokverk”. As we shall see in more detail below, none of these

2 Published by Den norske Bokklubben together with Gisla Saga Súrssonar and Laxdœla Saga in one volume under the common title Saga: norrøne sagaer i utvalg. 3 Aasmundsstad had used Konráður Gislason and Eiríkur Jónsson’s edition from 1875–89 whereas Liestøl could use Finnur Jónsson’s edition in the series Altnordische Sagabibliothek from 1908 (cf. bibliography).

45  On Translating Icelandic Sagas into Modern Norwegian versions in “Landsmål” / “Nynorsk” were complete. For this reason a new edition with the missing parts filled in was commissioned and published in 1996 and again in 2003. Even if these versions claim to be translations of the entire text4 they are not. Due to editorial inaccuracy, the publishers (Det Norske Samlaget) have, incredibly enough, failed to fill in missing genealogical information in seven chapters of the saga.5 The latter version was republished this year, in 2007, by Den Norske Bokklubben. This publication and the one from 1999 (Paasche’s translation) are, it seems, versions of Njáls saga widely distributed in present-day Norway (an impression of about 4000 copies each according to the publisher). Translations of Icelandic sagas in general should rely on the best editions published in the source language available at any moment. This is so, we must assume, when the translations of Brennu Njál’s saga into Norwegian are concerned. I say “we must assume” because some of them do not make this point explicit. The textual variation that can be observed between the various translated versions cannot, however, be accounted for, just by assuming that different editions of the source language text have been used. Here is not the time to go into detail about philological problems concerning the editing of a reliable text of the saga in its original language. As the editions upon which the various translations are based do not vary dramatically, we will just for the sake of convenience use the edition in the series “Íslenzk fornrit” (1954) as a point of reference in the following when comparing the translated texts, the overall impression of which displays textual variation between the different versions to a degree that may, somehow, remind us of the medieval manuscript transmission of saga texts.

4 “Ny og fullstendig utgåve ved Jan Ragnar Hagland” (‘New and complete edition by JRH’). 5 Chapters 20, 25 (two sequences), 26, 46, 95, 114, and 138. This is more than unfortunate as the preface to this version underscores the importance of completeness on this particular point in the narrative: “Utelating av slike delar av forteljinga kan difor både ta bort kulturhistorisk informasjon (jf. note 6 til kapittel 19 i soga) og fjerna litterære verkemiddel som skulle vera med og gje samanheng i teksten” (2003, s. 21). It is, for instance, not just a trivial detail to omit Ragnar Lodbrok from the genealogy of Snorri goði in ch. 114 and so on.

46  Jan Ragnar Hagland It seems fair to say, then, that the Brennu Njáls saga offered to the Norwegian reading public in modern and post-modern times varies to the extent that it is, in a certain sense, possible to see them as a number of different Njáls sagas. We will in the following try to look closer at the nature of textual differences that can be observed between the translated versions sketched above. The most important point of difference, it seems fair to say, is caused by abbreviation – or excision in Gérard Genette’s terms6 – the leaving out of various parts of the text in the source language, whatever edition has been used as basis for the translation. This, more than anything else, should in my opinion, be traced back to the translators’ ideas about translation and what a saga such as this should look like. As from Sommerfelt onwards all translations for a long time to some extent abbreviated the text on various points. Apart from Hallvard Lie’s 1941 version this is the case when all the translations into “Riksmål” / “Bokmål” are concerned and, unfortunately, also the ones into “Landsmål” / “Nynorsk”, the 1996–2007 versions included – even if the omissions there are due to editorial inaccuracy rather than deliberate choices made by the translator. The most important points of difference between the versions translated into Norwegian relate to the following aspects of the text: a. skaldic verse. b. introductory genealogical information. c. legal procedure. d. supernatural phenomena. We will take a quick look at each of these features in order to see how they are treated in the existing versions of Brennu Njáls Saga translated into Norwegian. The difficulties involved when trying to translate skaldic verse into any language are, to say the least, considerable. Brennu Njáls saga represents no exception in this respect. The translated versions of this saga into modern Norwegian have, as a whole, solved these problems in a variety of ways so as to create considerable variation between them. Again the most striking point of variation is created by the various omissions. As for the translation itself of the stanzas quoted in the saga, there is also much to be said. This is, however, a huge and general problem, that might fill a conference of its own, so we shall leave it at

6 See Genette 1997, 229f.

47  On Translating Icelandic Sagas into Modern Norwegian that for the moment. For the present purpose we shall just ask to what extent the translators have chosen to include the stanzas in the different translated versions. A total of 23 stanzas or parts of stanzas are quoted in the source language text – in the ÍF and other editions of the saga – in addition to the 11 stanzas of Darraðarljóð.7 Paasche’s 1922 version, Hallvard Lie’s 1941 version and Hagland’s 1996-2007 versions offer translations of all the stanzas. Paasche’s 1922 version, however, turns the verses in ch.12 into a prose line in direct speech. As can be expected the actual translations offered in the different versions vary a great deal, but that is a topic for another day. In all the remaining versions of the saga translated into modern Norwegian stanzas are, to a varying degree, omitted – most extensively so in Sommerfelt 1871, who omits 20 of the stanzas. Aasmundsstad was an autodidact in the Old Norse language and obviously reluctant to take responsibility for the difficult task of making sense of the stanzas. Therefore the linguist Rasmus Flo was commissioned to take care of that particular aspect of the saga text (Kleiven 1926, 25). Compared to Finnur Jónson’s 1908-edition and the ÍF edition of the source language text Aasmundstad’s 1896/97 version leaves out 12 stanzas and parts of stanzas. This version, however, in accordance with the 1875 edition used as the basis for the translation adds one stanza to ch. 7 and two to ch. 23 not extant in the reference texts mentioned. In Knut Liestøl’s 1928 version the stanzas included in the narrative follow Rasmus Flo’s selection in the 1896/97 version closely except for the three “extra” stanzas in ch. 7 and 23 which are omitted. A total of 12 stanzas or parts of stanzas are in consequence left out. The translation of each stanza is kept identical with or very close to the 1896/97 version. As in Paasche’s 1922 version the part of the stanza in ch. 12 is given in prose in these two early translations into “Landsmål”. In Aslak Liestøl’s 1961 version some of the stanzas are completely retranslated, some only slightly revised linguistically. By omitting

7 These are the  stanzas common to all complete manuscripts of the  saga. In the earliest complete manuscripts (e. g. Reykjabók, AM 468, 4to and Kálflækjarbók, AM 133 fol.) thirty so-called additional stanzas are included, stanzas that are included in some of the earlier editions (cf. Nordal 2007, 221 and 231f with references).

48  Jan Ragnar Hagland one stanza in ch.  132 and including three new ones in ch.  77, 145 and 157 the total number of stanzas omitted in this revised version amounts to 10. The versions published by Hallvard Lie in 1951 and 1954 are interesting in the sense that the translator himself abbreviates by omitting several of the stanzas from his own complete version of 1941 – most extensively so in the 1951 version in which 11 stanzas are omitted. In the 1954 version four stanzas are left out. Hallvard Lie does not make a specific point of these changes in the preface to these versions (almost identical in the two). As Njáls Saga is the longest of all the Icelandic family sagas, he says in the prefaces, abbreviations have been made – abbreviations of which he makes specific reference only to those concerning genealogy and legal procedure (point b and c above). The omission of skaldic verse has, however, wider consequences for the translations than the text constituted by the omitted stanzas as such. This is so because even sequences of narrative prose surrounding the stanzas have frequently been suppressed in order to “conceal” the abbreviation. It is worth noticing, I think, that the 1961 version in “Nynorsk” except for one single stanza omits the very same ones as does the version in “Bokmål” of 1951. Even if it cannot easily be proved there is no reason to believe that this is just due to chance. It seems fair to say, then, that the translators’ attitudes concerning the importance of skaldic verse in the saga alone have created a variety of Njáls Sagas offered to the reading public in Norway over the years. This variety has been deepened further by similar attitudes towards the importance of genealogical information and the depiction of legal procedure in the text. For the present purpose I shall, in order to illustrate the two latter points, just quote from Hallvard Lie’s preface to the 1954 version along with a footnote made by Fredrik Paasche to the 1922 version in order to justify a major omission in ch. 142 of the saga. Hallvard Lie justifies his abbreviations as follows: ‘It is particularly the long and – for modern readers – tedious pleadings of the Allthing scenes that have been affected. Also the genealogies have been shortened, as these – for a saga reader of to day – are totally “dead matter”; for the old Icelanders, however, these were exquisite literary delicacies. Otherwise minor

49  On Translating Icelandic Sagas into Modern Norwegian abbreviations have been made here and there where possible without weakening the general artistic effect in any way’.8 Paasche in his footnote to the omitted sequence of ch. 142 states that ‘Here for the first time some of the text has been omitted. The omitted piece deals with Mørdr’s conduct of the case, which is quite prolix. What Mørdr does can be deduced from his own ensuing words’.9 These quotes unveil, I think it is fair to say, a “readers’ digest” kind of attitude, very noticeable in the mediating of Brennu Njáls Saga by way of translation into modern Norwegian over the years. We shall not expand on that here, suffice it so say that Norwegian speaking students – if dependant on translations – should, in consequence, be very careful when choosing a translated version of this particular saga. A final point about abbreviations to be made here relates to what could be termed supernatural phenomena in the text – point d above. Towards the end of the saga, in chapters 156 and 157, three sequences, two of them quite long, have been omitted in some of the translated versions. They all tell about miracles – jarteikn – of the kind often found in legendary texts. Sommerfelt 1871, Paasche 1922, and Lie 1941 do not make these omissions.10 Hallvard Lie’s 1951 version leaves them all out, whereas only one of the two sequences in ch.157 is left out in the 1954 version, so as to make a rather strange pattern. Among the translations into “Landsmål” / “Nynorsk” these abbreviations exist in the 1961 version only. Probably this version just copies Hallvard Lie’s abbreviation

8 “Det er især de lange og  – for moderne lesere  – nokså trettende prosessinnleggene i tingscenene det er gått ut over. Slektsregistrene er også blitt beskåret, da de for en norsk sagaleser i dag er totalt «dødt stoff»; for de gamle islendinger var de derimot en utsøkt litterær lekkerbisken. Her og der ellers er også mindre forkortninger foretatt, hvor det kunne skje uten at sagaens kunstneriske helhetsvirkning på noen måte ble svekket” (Lie 1954, 10). 9 “Her er det for første gang noe av teksten blir utelatt. Stykket handler om Mørds saksfremlegg og er meget vidløftig. Hva Mørd foretar seg, fremgår av hans egne ord i det følgende” (1986 reprint p. 258). 10 The parts left out are a long sequence telling about the raining of blood, swords fighting by themselves, and an attack by ravens with iron claws in ch. 156 and two shorter sequences in ch. 157, one about the healing of the boy Taðkr and one about various supernatural events happening in the Faroe islands, in Iceland and in the Orkneys (cf. Hagland 1987, 47f.).

50  Jan Ragnar Hagland from 1951 on this point. It is worth noticing that no reasons for omitting these parts of the narrative are given in the versions that do so. We might suspect that narrative elements such as these did not satisfy Hallvard Lie and Aslak Liestøl’s ideas of what a realistic saga like this ought to look like – that these supernatural events did not belong there. Hallvard Lie’s inconsistency on this point in his versions, however, makes it difficult to understand the textual variation he creates on this point. As far as the historical reception in general of Brennu Njáls Saga in Norway is concerned, then, it seems fair to say that one translated version or other of the saga has been available to the reading public more or less continuously from the 1870s onwards. In statistical terms it seems as if the most abbreviated version – the 1951 version – had the widest distribution. It is not irrelevant, then, as we have seen, to ask what Njáls saga we are referring to when speaking about its historical reception in Norway. The quantitative aspects of this do not, however, lead us very far. On the qualitative level we have unique and interesting information about the importance of Njáls Saga – even for the Norwegian history of literature. Well known are Sigrid Undset’s own words about the importance of her first encounter with this particular saga. In an essay called “A book that was a turning point in my life”11 she reflects upon this encounter, of which she had given details already in her strongly autobiographical novel Eleven years (Undset 1934). In our context it is relevant and interesting to note that it was Sommerfelt’s version from 1871 that made such a decisive and lasting impression on the 11 years old girl who was later to become Nobel Prize laureate in literature. We do not, to my knowledge, have information that can stand up to this when it comes to the reception of other translated versions of this saga into Norwegian. Sometimes the history of translation concerning a text such as Brennu Njáls Saga may even reflect changes of attitudes in the society at large surrounding the texts transferred into the target language. The famous episode in ch. 7 of the saga – where Unnr reveals to her father, Mörðr, her reasons for wanting a divorce – may serve as a nice 11 Printed posthumously in 1952, translated from the English – “En bok som blev et vendepunkt i mitt liv” (Undset 1951, 27–34).

51  On Translating Icelandic Sagas into Modern Norwegian little example of this. The editions of the saga in the original language – as from the 1772 edition onwards – quote the reasons she gives for this quite straightforwardly without evasion. For the 19th century translators this obviously was a difficulty that had to be solved by paraphrase in more or less euphemistic terms. We see this clearly in Sommerfelt’s version. His translation, it seems, copies N. M. Petersen’s rather bashful solution in his version from 1841 and 1862 on this particular point, it is at least very close. This is also the case with Sir George Webbe Dasent’s translation into English from 1861 – a tradition that was continued in the Norwegian versions of 1896/97 and 1928. Fredrik Paasche in his 1922 version was the first to translate this part of the text into Norwegian without paraphrasing it. Sir George Webbe Dasent’s somewhat timid translation into English may well represent the bashfulness that also Norwegian translators of the 19th and early 20th centuries experienced when trying to mediate this particular point in the text: “when Mord pressed her to speak out, she told him how she and Hrut could not live together, because he was spell-bound, and that she wished to leave him”.12 Why give the indelicate details of the source language text when they can be avoided so elegantly – we may well ask! Or perhaps prudishness ought to be included among the causes for variation in translation? 12 Cf. e. g. the 1772 edition of the saga on this point: “Hversv má svá vera? segir Morðr. ok seg enn giorr. hon svarar. þegar hann kemr við mik þá er horvnd hans svá mikit at hann má ekki eptir lęti hafa við mik. en þó hofvm við bęði breytni til þess á alla vega at við męttim niótaz. en þat verðr ekki. en þó aðr við skilim sýnir hann þat af ser at hann er í ęði sínu rett sem aðrir menn.”

Bibliography Editions: [Ólafur Olavius ed.] 1772, Sagann af Niáli Þorgeirssyni ok Sonvm hans &c. út gefin efter gavmlum Skinnbókvm med Konvnglegu Leyfi. Kaupmannahafn. [Konráður Gíslason & Eiríkur Jónsson (eds.)] 1875–89, Njála. Udgivet efter gamle håndskrifter af det Nordiske Oldskrift-Selskab. København. Finnur Jónsson (ed.) 1908, Brennu-Njálssaga (Njála) = Altnordische Sagabibliothet 13. Halle: Verlag von Max Niemeyer.

52  Jan Ragnar Hagland Einar Ól. Sveinsson (ed.) 1954, Brennu-Njáls Saga = Íslenzk fornrit, vol. XII. Reyk­ javík: Hið íslenzka fornritafélag. Translated versions (listed in order of chronology): 1861 = The story of burnt Njal, or Life in Iceland at the end of the tenth century. From the Icelandic of the Njals saga by George Webbe Dasent. Edinburgh: Edmonston and Douglas. 1862 = Njals saga: eller fort lling om Njal og hans sønner. Efter de islandske grundskrifter ved N. M. Petersen. Anden udgave. København 1962. [1st ed. 1841 = Historiske Fort llinger om Isl ndernes F rd hjemme og ude Udg. af det Kongelige Nordiske Oldskrift-selskab ved N. M. Petersen. Vol 3.] 1871 = Njaals Saga oversat af Karl L[inné]. Sommerfelt. Kristiania: Selskabet for Folkeoplysningens Fremme. 1896–97 = Njaala elder Soga um Njaal Torgeirsson og sønerne hans. Umsett fraa gamalnorsk av Olav Aasmundstad. Kristiania: Det Norske Samlaget. 1922 = Njaals saga. Oversat av Fredrik Paasche. Kristiania: H. Aschehoug. Reprints 1986, 1997. Thorleif Dahls Kulturbibliotek. Oslo: H. Aschehoug. 1999 In: Saga: Norrøne sagaer i utvalg. Oslo: Den norske Bokklubben. 1928 = Njåla umsett av Olav Aasmundsstad. Ny utg. ved Knut Liestøl = Norrøne Bokverk 23. Oslo: Det Norske Samlaget. 1941 = Njåls Saga. Oversatt av Hallvard Lie. Oslo: Gyldendal Norsk Forlag. 1951 = Njåls Saga. Oversatt av Hallvard Lie = Hjemmenes boksamling. Oslo: Gyldendal Norsk Forlag. 1954 = Njåls Saga. Oversatt av Hallvard Lie. In: Islandske ttesagaer. Under redaksjon av Hallvard Lie, vol. IV, pp. 7–252. Oslo: H. Aschehoug & Co. 1961 = Njåla. Til nynorsk ved Aslak Liestøl. In: Den norrøne litteraturen, bd. 2. Ættesoger. Oslo: Det Norske Samlaget. Reprinted in the series ‘Norrøne bokverk’, No 23. Oslo 1975: Det Norske Samlaget. 1996 = Njålssoga. Omsett av Aslak Liestøl. Ny og fullstendig utgåve ved Jan Ragnar Hagland. Oslo: Det Norske Samlaget. 2003 = Njålssoga. Innleiing ved Roy Jacobsen. Omsett av Aslak Liestøl. Ny og fullstendig utgåve ved Jan Ragnar Hagland. Oslo: Det Norske Samlaget. 2007 = Njålssoga. Oversatt av Aslak Liestøl. Ny og fullstendig utgave ved Jan Ragnar Hagland. Innledende essay av Roy Jacobsen. Oslo: Den norske Bokklubben. Other references Genette, Gérard 1997 [1982], Palimpsests. Literature in the Second Degree. Lincoln and London: Nebraska University Press. Original title 1982: Palimpsests: La literature au second degree. Paris: Seuil. Guðrún Nordal 2007, The Art of Poetry and the Sagas of Icelanders. In: Quinn, Judy & al (eds.) Learning and Understanding in the Old Norse World. Essays in Honour of Margaret Clunies Ross, pp. 219–237. Turnhout, Belgium: Brepols Publisers n. v.

53  On Translating Icelandic Sagas into Modern Norwegian Hagland, Jan Ragnar 1987, Njåls saga i 1970- og 1980-åra. Eit oversyn over nyare forskning. Scripta Islandica. Isländska sällskapets årsbok 38, pp. 36–50. Hagland, Jan Ragnar 1996, “Fager er lidi”. Eit hundreårsminne for Olav Aasmundstads Njålssoge på landsmålet. Motskrift No 2 1996, pp. 157–169. Hagland, Jan Ragnar 2003, Å meisla ut eit litteraturspråk. Kring litterære omsetjingar til Landsmaalet fram til om lag 1900. Motskrift No 1 2003, pp. 43–55. Jón Karl Helgason 1999, The Rewriting of Njáls Saga. Translation, Ideology and Icelandic Sagas = Topics in Translation 16. Clevedon–Buffalo–Toronto-Sydney: Multilingual Matters Ltd. Kleiven, Ivar 1926. Olav Aasmundstad. Bygd og Bonde. Tidsskrift for historie og folkeminne. 8. årg., pp. 19–27 Undset, Sigrid 1934, Elleve aar. Oslo: H. Aschehoug & Co. Undset, Sigrid 1952, Artikler og taler fra krigstiden. Utg. av A. H. Winsnes. Oslo: H. Aschehoug & Co.

Norse Kings’ Sagas Spread to the World Jon Gunnar Jørgensen The University of Oslo

1. The writing of Kings’ sagas The Kings’ sagas are regarded as a separate genre of Icelandic saga literature. They have been, and still are of extraordinary value as sources of Scandinavian history, but also from an aesthetic point of view – as literature. The Kings’ sagas were also the first part of the Icelandic saga treasure to awaken the attention abroad in the 16th and 17th centuries. The main subject of these sagas is the lives and political career of Scandinavian, mainly Norwegian, kings from the Viking Age and through the life of King Hakon Hakonsson (d. 1263). It is a mystery how and why this literature developed in the periphery area of Iceland, at a time Icelanders did not even have a king of their own. It is hard to believe that pure historical interest could have been sufficient motivation. Perhaps it was a way to discuss the problematic lack of a Head of State which was prescribed by the medieval models of government. This could be actualised by the serious internal conflicts of the Sturlunga era. The writing of sagas could also be motivated by a request for a cultural capital, as some Icelandic scholars have argued recently, with the reference to the ideas of the French sociologist Pierre Bourdieu (See e. g. Torfi H. Tulinius 2004). Saga manuscripts have also been pointed out as a valuable article of export (see Stefán Karlsson 1979). The Icelandic Sagacodices have no doubt been of great value to the Norwegian aristocracy and the kings of Norway. However, this cannot explain the developing of the art in Iceland, only that several saga codices found their way to Norway in the late Middle Ages. Whatever the explanation can be, it is a fact that a literature of outstanding value about Scandinavian – especially Norwegian – history was produced in Iceland in the 13th century. In the following centuries, these sagas caught the attention of scholars and readers in all Europe and the whole world. Approaching the Viking Age. Proceedings of the International Conference on Old Norse Literature, Mythology, Culture, Social Life and Language. Edited by Ērika Sausverde and Ieva Steponavičiūtė. (Scandinavistica Vilnensis 2). Vilnius University Press, 2009. https://doi.org/10.15388/ScandinavisticaVilnensis.2009.2.4 Copyright © 2019 Authors. Published by Vilnius University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

56  Jon Gunnar Jørgensen The  golden age of the  Kings’ sagas was contemporary with the developing of the medieval kingdom of Norway, beginning with king Sverrir and reaching its zenith under Hakon Hakonsson. Snorri Sturluson was an older contemporary of Hakon, and also closely connected to him. He visited Norway twice during Hakon’s rule. After the decay of the Norwegian Realm the composing of Kings’ sagas also diminished. It seems to have been some activity of reproducing sagas in great codices through the first decades of the 14th century, but then this interest also seems to have decreased. The Flatey-book was written at the end of the 14th century, and can be seen as a grand finale for the genre in Iceland. In the 15th century, the romantic sagas take over the scene. In the 14th century the Norwegian Realm fell apart. As the court moved east to Sweden with Magnus Eiriksson in the first half of the century, the Norwegian aristocracy disbanded, and so did the interest for the sagas. Seen from the 15th century viewpoint, the writing and reading of Kings’ sagas belonged to the past. The beautiful vellums still were on Norwegian shelves, but nobody paid any attention to them. And why should the Icelanders bother? Why should they read and rewrite these sagas? They did not. The Kings’ sagas now belonged to the past. They had in double meaning become history – they were no longer relevant to the present. But at one point, the sagas were somehow brought back to the stage. I will look into how this re-introduction of the Kings’ sagas happened, and present the first three printed editions. The first two introduced the material to Scandinavians readers and historians. I will here, however, emphasize the third one, from 1697 that made the sagas available to the whole of learned Europe. The three editions are: Norske Kongers Krønicke oc bedrifft. Copenhagen 1594 ( Jens Mortensen’s Chronicle) Norske Kongers Chronica. København. Copenhagen 1633 (Peder Claussøn Friis’ Saga Translations) Snorre Sturlusons Nordländske Konunga Sagor. Stockholm 1697 [–1700] ( Johan Peringskiöld’s Edition)

57  Norse Kings’ Sagas Spread to the World 2. The revival of the sagas In the 16th century, a new interest for historiography developed in Europe, and one turned again to the medieval sagas. This new attitude was introduced in the continent by the renaissance humanists. History could not only quench learned peoples thirst for knowledge, it could also be a chessman in the game of power. Along with the rivalry between royal authorities and the church, the historical chronicles could be used to strengthen the king’s power and to legitimate his rule, in a similar way to how the church used literature in its propaganda. The academic renaissance for the classic profane literature, written in Latin, had grown strong in Europe in the 15th century, and at the beginning of the 16th, it also reached Scandinavia. A new quest for historical literature followed. I consider it probable that the publishing of Saxo Grammaticus’ Gesta Danorum, the Latin chronicle of the Danes, in 1514, may have been the trigger for the new reading of the Norse sagas. This edition was widely spread and became a most prestigious work. The reading of Gesta Danorum most often just called “Saxo” also inspired small learned circles in Norway, first of all in Bergen. After having published “Saxo”, the editor, Christiern Pedersen, caught interest in the historical material hidden in old, half-forgotten manuscripts in Norway. Somehow, he must have been informed of their existence, and at a time, probably in the 1530’s, he requested some excerpts of the sagas to be translated into Danish. We are not certain who wrote these excerpts, but there are indications that it might have been the later magistrate (Norw.: lagmann) of Bergen, Jon Simonssøn (1512–1575) ( Jørgensen, Jon Gunnar 1993/2, esp. pp. 169–171 and 182–186). The socalled “Christiern Pedersen’s Excerpts” are the first signs of the revived interest in Kings’ sagas after the Middle Ages. Members of the learned circles of Bergen did several approaches to the Saga literature in the 16th century. In addition to “Christiern Pedersen’s Excerpts” at least two other translations – or, more correctly, paraphrases of Kings’ sagas in Danish were made and sent to Copenhagen. One of them, that goes by the name “Laurents Hanssøn’s Saga-translation” was made for a definite purpose; namely, for the education of the Danish crown prince Frederik. The text is only found in one single manuscript, which still exists in the Arnamagnæan Collection in

1. The portrait of Peder Claussøn Fries in Valle Church, near Lindesnes on the southern edge of Norway. Some biographical keynotes are written in the left upper corner, but the year of his death is written later on the opposite side.

Copenhagen (AM 93 fol.). The other translation, made by the magistrate of Bergen, Mattis Størssøn, must have been spread among interested colleagues, because it is known from several manuscripts. One of these fell into the hands of the historian Christopher Huitfeldt in Copenhagen, who engaged Rev. Jens Mortensen to have it published. Norske Kongers Krønicke oc bedrifft was printed in 1594. It was a small book, the text was both corrupted and shortened, but it was very important. The translator, Mattis Størssøn, is not mentioned at all; probably he was not even known by the editor. Neither can we find one word about Snorri Sturluson nor the Icelandic origin of the text. That kind of information we could not either expect, since it was unknown at the time. The publishing of this book is remarkable, because it is the first printing of Old Norse saga material at all. Now the door to the saga literature was opened for readers and scholars outside the narrow circles in Bergen and Copenhagen. Though the  text of this first edition was short and paraphrasing, Ynglinga saga was reproduced in a fuller shape. Ynglinga saga

59  Norse Kings’ Sagas Spread to the World is the first part of Heimskringla, containing the mythic pre-history of the Norwegian kings’ kin. This material was a special treasure to the Swedes. I will soon return to the enthusiastic reception of the sagas in Sweden. The 1594 chronicle was the pioneer edition of Kings’ Sagas, and it was printed in a very limited number of copies. Nobody could predict the formidable demand for saga literature in the future. That explains why the edition is so rare today; it is only to be found in very few academic libraries and in a few private collections, in all probably not more than ten copies.1 The 1594 edition was important as a first introduction to the material. Besides its great importance to the Swedes, it probably also inspired and motivated the next edition, published in 1633 (also in Copenhagen). 3. The Scandinavian break-through for Kings’ sagas Shortly after the 1594 edition was published, Peder Claussøn Friis (1545–1616) started to prepare a far more extensive one. Peder Claussøn was a vicar in Audnedal, on the southern edge of Norway. The key information of his biography is painted on his portrait, situated in Valle Church, where he practised. He may have looked like this. The portrait is probably painted while he was still alive, since the year of his death was added later. Fifteen years after Peder’s death, the famous Danish scholar Ole Worm became acquainted with the translation. He obtained a manuscript and published it in 1633. This was the first holistic printed reproduction of the sagas. This edition was certainly printed in a larger number of copies, since it still is rather easy to find, both in libraries, private collections and at antiquarian booksellers, and it had tremendous influence. Due to this very edition Heimskringla was established as a monument in the Scandinavian history of literature. Today the Icelandic “Snorre” (i. e. Heimskringla) is considered at the top of the canon of literature in Norway, above the works of Norwegian authors. The position of Heimskringla in Scandinavia and later especially in Norway was

1 According to unverified information from Deichmanske bibliotek, the Munici­ pal Library of Oslo, there are six in public libraries. Two are known in private collections. Not all of the copies are complete.

2. The title page of Peder Claussøn’s saga translation, published in Copenhagen 1633 by Ole Worm. It tells us that Snorre Sturlason is the author of the work. The vignette shows the Norwegian Coat of Arms, “The Norwegian Lion”.

established by the 1633-edition, and also, by this edition, the author Snorri Sturluson was connected to the work. That was news in the 17th century Scandinavia, and also in Iceland. Peder Claussøn’s Norske Kongers Chronica (NKC) became important in all the Scandinavian countries, including Iceland. I have already mentioned that the attribution to Snorri was unknown also in Iceland till the book was published, even by the most learned specialist, Arngrímur Jónsson (1568–1648). The acquaintance with the book also led to a revival of interest in the Kings’ sagas in Iceland. Now Icelandic scribes began to reproduce Kings’ sagas anew. Through lack of old Icelandic manuscript exemplars, some scribes took the text from Peder Claussøn and translated parts of it back into Icelandic. In Sweden, the edition became an important source for Swedish historians and antiquarians. Gothicists believed that Goths were the ancestors of Swedes, and that the Old Norse language was Gothic.

61  Norse Kings’ Sagas Spread to the World Therefore, the Old Icelandic literature in their eyes also had its roots in Sweden. The NKC 1633 served as inspiration, source and editorial model for the first Swedish edition (1697), which I will soon present. In Denmark, NKC found its place besides “Saxo”. It is evident that the two works were soon regarded as a couple – the twin-chronicles of the twin-states. If you look at the vignette on the title page, you see the Norwegian Lion, the national coat of arms. Further the book is dedicated to Christopher Urne, who was the governor of Norway in 1633. After 1633 and till the end of the 19th century, new translations of “Saxo” were always planned together with new or revised translations of “Snorre”. In 1757 a revised version was published in Copenhagen. The text had been completely rewritten, but the editor was not even mentioned on the title page. All tribute went to the great Peder Claussøn, and when the monumental scholarly edition of Kings’ sagas was being prepared in Copenhagen starting in 1770’s and under the leadership of Gerhard Schønning, the editors were instructed to keep Peder Claussøn’s style in the new translation. It was doubtless the 1633 edition that established the position of the Kings’ sagas also in the Norwegian, National canon. But it did not happen until the 19th century, when the new independent nation developed, that “Snorre” really grew big in Norway. Now the thread was taken up from Peder Clausson and new translations made. These have found their way literally into every home. 4. Peringskiöld 1697. From Scandinavia to the wide world We have now followed the Kings’ sagas on their journey from Iceland to Norway, and further to Denmark. Through the two first printed editions, the sagas spread in the Nordic countries. But still, because of the language, they were not introduced to the learned European community. The discovery of Ynglinga saga was especially celebrated with enthusiasm in Sweden. Swedes had ever since 1514 envied their Danish neighbours and rivals their “Saxo”. In the following decades, Swedish authors made several attempts to write a similar proper history of Swedes. Best known is the one by Johannes Magnus, Historia de omnibus gothorum sueonumque regibus (‘History of all Kings of Goths and Swedes’), printed in 1554. But the Swedish historians had a serious problem – the lack of written sources. In Ynglinga saga they could read

3. The first edition of the Icelandic text of Heimskringla was published in Stockholm, 1697, edited by Johan Peringskiöld. The ­edition was provided with translations both to Swedish (in the right column) and Latin (at the bottom).

that Odin and his companions, the “diar” settled in Uppsala, and made Sweden the home of the Ynglings, the ancestors of the Norwegian kings. That was really a fat bite for the Gothicists. In the middle of the 17th century a strong historical academy was founded in Uppsala, Collegium Antiquitatum (‘College of Antiquities’). A large community of scholars and assistants did historical studies and published historical sources and writings. In 1679 the Icelandic student Guðmundur Ólafsson was engaged as the successor of his countryman Jonas Rugman. Now a prestigious edition of Heimskringla received high priority. It was Johan Peringskiöld (the elder) who became the leading editor and person-in-charge of the project and Guðmundur Ólafsson, who had been recruited at the University of Copenhagen, became his indispensable assistant, helping his superior to understand the Icelandic text and translating it into Swedish. Another Icelander, Jón Eggertsson, occasionally staying in Copenhagen, transcribed for them the most important Heimskringla manuscript – “Kringla”. He also made a journey to Iceland in order to collect and buy manuscripts for the Swedes,

63  Norse Kings’ Sagas Spread to the World a task he accomplished with great success, though it was against the law. The Danish king had forbidden Icelanders to transmit manuscripts to any other than his own representatives. One of the manuscripts that Jón bought for his Swedish employers was the so-called Húsafellsbók, a 17th century manuscript containing Kings’ sagas, compiled with NKC 1633 as a template. Large parts of the text were translated from the Danish printed text and back to Icelandic (see Ólafur Halldórsson 1976). As copy-text for his edition Peringskiöld used the “Kringla” transcription by Jón Eggertsson. But in addition he used every other available source. The whole “Kringla” text was collated against NKC and Húsafellsbók, and if a passage or some information should be missing in “Kringla”, it was carefully supplied by the other sources. The additional text from Húsafellsbók followed Heimskringla through later editions for more than a century. Peringskiöld’s edition was a great step forwards as a source publication. It reproduced the text from the manuscripts far more confidently than the older Danish editions, and for the first time the Heimskringla text was printed in the original language. But the edition offered more – it was trilingual. In addition to Icelandic, the text was translated both into Swedish and Latin. The concept with a synoptic presentation of the Icelandic text with a Swedish translation had been used in several editions from the Collegium Antiquitatum before Heimskringla, and this was also the intention now. The Swedish translation was done by Guðmundur Ólafsson and revised by Peringskiöld. But as the work took form, Peringskiöld applied to the chancellery for time and funding to provide the edition also with a Latin translation, so that the whole world of learned could share this treasure, and be acquainted with the ancestors of the Goths. The authorities were favourable, and the Latin translation was produced by Peringskiöld himself. The result, especially the Latin translation, was naturally heavily criticised by Danish scholars – they were certainly envious. Nevertheless, the work was really impressive as a whole. The text covers the whole Heimskringla, from Ynglings to Magnus Erlingsson in two volumes. The first volume was published in 1697, the second a couple of years later, probably 1700. The title Heimskringla was in this edition related to the work. Earlier it had been used with reference to the manuscript “Kringla”, the copy‑text for the edition.

64  Jon Gunnar Jørgensen My first thought when I examined Peringskiöld’s edition, was that the prior purpose of the Latin text was to give the volumes an academic impression and prestige. I questioned whether the book really had a market outside the Nordic countries. It might, of course, be a matter of importance that “Saxo” was in Latin, and therefore desirable also to have Heimskringla available in Latin. It might be that Latin primarily simply signalized that the content was of scholarly significance. It seems, however, that the edition really found its way to the learned Europe and that there definitely was an international interest. It was reviewed in foreign magazines and referred to by learned writers. As early as in 1698, Lorentz Hertel refers to the edition in a letter to Gottfried W. Leibniz (Leibniz 1998, pp. 48–49). Then later in the 18th century we see that the French and English really became interested in Old Norse culture, especially after Paul-Henry Mallet’s history of the Danes2 and other writings about Scandinavia. Not the least due to Mallet, Icelandic sagas were introduced to the French during the age of Enlightenment. In the following century of Romanticism saga literature really rooted in the European literary reception. The three editions presented here were important steps to clear the way for the treasure of Icelandic sagas into the European canon. References Torfi H. Tulinius. “Kapital, felt, illusio”. Maal og Minne 2004, pp. 1–20. Stefán Karlsson. “Islandsk bogeksport til Norge i middelalderen”. Maal og Minne 1979, pp. 1–17. Jørgensen, Jon Gunnar. “Sagaoversettelser i Norge på 1500-tallet”. Collegium Medie­ vale 1993/2, pp. 169–198. Ólafur Halldórsson. “Um Húsafellsbók”. Minjar og menntir: Afm lisrit helgað Kristjáni Eldjárn 6. desember 1976. Reykjavík 1976, pp. 391–406. (Reprinted in Grettisf rsla. Reykjavík 1990). Leibniz, Gottfried Wilhelm. Sämtliche Schriften und Briefe. Erste Reihe: Allgemeiner politischer und historischer Briefwechsel. Berlin 1998.



2 Introduction à l’Histoire de Dannemarc (Introduction to the History of Den­ mark). 1 vol. first published 1755.

Sámi and Scandinavians in the Viking Age Jurij K . Kusmenko Institute of Linguistic Studies, St.-Petersburg, Russian Academy of Sciences

Introduction Though we do not know exactly when Scandinavian and Sámi contact started, it is clear that in the time of the formation of the Scandinavian heathen culture and of the Scandinavian languages, the Scandinavians and the Sámi were neighbors. Archaeologists and historians continue to argue about the location of the original southern border of the Sámi on the Scandinavian peninsula and the location of the most narrow cultural contact, but nobody doubts, that the cultural contact between the Sámi and the Scandinavians before and during the Viking Age was very close. Such close contact could not but have left traces in the Sámi culture and in the Sámi languages. This influence concerned not only material culture but even folklore and religion, especially in the area of the Southern Sámi. We find here even names of gods borrowed from the Scandinavian tradition. Swedish and Norwegian missionaries mentioned such Southern Sámi gods as Radien (cf. norw., sw. rå, rådare), Veralden Olmai (< Veraldar goð, Freyr), Ruona (Rana) (< Rán), Horagalles (< Þórkarl), Ruotta (Rota). In Lule Sámi we find no Scandinavian gods but Scandinavian names of gods such as Storjunkare (big ruler) and Lilljunkare (small ruler). In the Sámi languages, we find about three thousand loanwords from the Scandinavian languages and many of them were borrowed in the common Scandinavian period (550–1050), that is before and during the Viking Age (Qvigstad 1893; Sammallahti 1998, 128–129). The famous Swedish lappologist Karl Bernhard Wiklund said in 1898 “[…] Lapska innehåller nämligen en mycket stor mängd låneord från de nordiska språken, av vilka låneord de äldsta ovillkorligen måste vara lånade redan i urnordisk tid, dvs under tiden före ca 700 år efter Kristus. Dessa urnordiska låneords mängd visar, att lapparna redan vid denna tid måste ha stått i en mycket intim beröring med skandinavierna, så intim, att de båda folken bör ha bott i hvarandras omedelbara närhet och icke endast kommit i beröring med hvarandra under Approaching the Viking Age. Proceedings of the International Conference on Old Norse Literature, Mythology, Culture, Social Life and Language. Edited by Ērika Sausverde and Ieva Steponavičiūtė. (Scandinavistica Vilnensis 2). Vilnius University Press, 2009. https://doi.org/10.15388/ScandinavisticaVilnensis.2009.2.5 Copyright © 2019 Authors. Published by Vilnius University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

66  Jurij K. Kusmenko några sällsyntare handelsresor e dyl.” (unpublished oral presentation in 1898, quoted after Fjellström 1985, 118). Nobody denies the fact of great Scandinavian influence on the Sámi in the Viking Age. But if we ask whether this Sámi-Scandinavian contact is reflected in Scandinavian culture, we get as a rule a negative answer. Scandinavian historians and linguists are, with very rare exceptions, unanimous in this case. The Scandinavian influence on the Sámi languages and on the Sámi culture on the one hand and stigmatization of the Sámi in the 19th and beginning of the 20th centuries, which has been extrapolated to the whole period of the Sámi-Scandinavian contacts on the other, has made an assumption about Sámi influence on the Scandinavians impossible. The proposed low social prestige of the Sámi and even their stigmatization determined the assumption about the influence only in one direction1. The traditional opinion at the end of the nineteenth and the beginning of the twentieth century was that everything in the Sámi religion was borrowed from the Scandinavians.2 Respectively, Sámi influence on the Scandinavian languages has been rejected.3 In the new very thick compendium on historical Scandinavian linguistics, where a special chapter is devoted to Scandinavian language contact, we shall look in vain for something about the Finnic-Ugric influence on the Scandinavian languages. The only information about the matter are the words of Koivulehto that such an influence “does not seem very probable” (Koivulehto 2002, 590–591). Thus the possibility of a Sámi influence on the Scandinavian culture and on the Scandinavian languages was rejected from the very beginning. But was it really so that the relation between the Sámi and the Scandinavians in the Viking Age was the same as it was at the end of the 19th and the beginning of the 20th century and excluded every Sámi cultural and language influence on

1 cf. “The cultural way was usually a one-way street from the Germanic people to the Finns or the Lapps” (Einarsson 1986, 43). 2 cf. “Det vil være ørkesløst at regne med en hjemmefødt lappisk kultur” (Olrik 1905, 44). 3 cf.  “Only few Finnish or Sámi loanwords concerning special Sámi and Finnish matters and some marginal features in the outmost northern Swedish and Norwegian dialects are the only possible Sámi or Finish influence on the Scandinavian languages” ( Jahr 1997, 943; cf. also Sköld 1961, 64).

67  Sámi and Scandinavians in the Viking Age the Scandinavians? In this paper I try to show that the cultural influence in the Viking Age was not one-sided. The cultural impulses went not only from the Scandinavians to the Sámi but even to a very strong degree from the Sámi to the Scandinavians. My sources will be archaeology, onomastics, Old Norse literature and Scandinavian languages. Archaeology The present Sámi territory stretches from Idre Parish in the Swedish province of Dalarna to the Kola Peninsula. However, during the Viking Age, the Sámi territory reached much further south than has been assumed up to now. The traditional point of view that the southern Sámi did not appear in central Norway and central Sweden until the 16th–17th centuries (cf. Sandnes 1973, Haarstad 1992) has been revised recently. The latest archaeological and historical studies give evidence of a Sámi population that possibly reached as far south as the Mälardal region in present central Sweden and eastern Norway, see the maps in Zachrisson 1997; 2004; 2006.4 Zachrisson assumes a large zone of Sámi-Scandinavian cultural contacts during the Viking Age in central Scandinavia including in Trøndelag, Oppland, Hedmark, Jämtland, Härjedalen, Ångermanland, Värmland, Dalarna, Medelpad, Hälsingland and parts of Buskerud, Telemark, Akerhus, Västmanland and Uppland. For this contact zone she proposes “en viss kulturell simbios” (Zachrisson 1997, 218). Hansen and Olsen write about “en ikke ubetydlig grad av kulturell kreolisering” in this area (Hansen, Olsen 2004, 107). But the connection between archaeological culture and ethnic identity and language is very problematic. That people wear jeans does not mean that they are Americans and speak English. We do not

4 cf. “Samene har vært i sørsamisk område langt tilbake i førkristen tid” (Salvesen, 1980, 147); The authors of the latest Sámi history Hansen and Olsen affirm: “der er lite grunnlag for å se den historisk kjente sørsamiske tilstedeværelsen som et resultat av “innvandring” fra nord. Det er tvert om grunn till å anta at det samiske bosetningsområdet i sør var langt større enn hva dagens situasjon og nyere historiske kilder antar” (Hansen, Olsen, 2004, 109). Cf. Also Sammallahti 1990, 441. Zachrisson’s map has not been accepted with unanimous approval of archaeologists, cf. Baudou 2002, 31. But even Baudou in his early book has drawn the movable border between two cultures in Gästrikland and along the Dalaälven (Baudou, 1995, 53), which is not so far from the southern border of Zachrisson’s contact area.

68  Jurij K. Kusmenko know what language was spoken by the representatives of the mixed Sámi-Scandinavian culture, or what folklore they had. Was it so that the Scandinavians adopted Sámi features or the Sámi took on Scandinavian cultural features and language? Moreover there is a tendency in the latest archaeology to deny a connection between archaeological culture and ethnic identity at all.5 Indeed, in some cases we cannot find a connection between material culture and ethnic identity as, for example, in the case of the Sámi, who identify themselves as Sámi, but do not speak the language, are not involved in reindeer herding and live in Stockholm. Many other cases show us the lack of an obligatory connection between material culture and ethnic identity. In many cases, however, this connection is obvious. That means that the archaeological data could be used only in connection with the data of spiritual culture and with the linguistic data. Fortunately, we have at our disposal not only archaeological sources but also other sources, particularly Old Norse literature and onomastics. The Sámi in Old Norse literature A very important source helping us to understand the role of the Sámi in the Scandinavian society is Old Norse literature6. The usual cliché features of Sámi are that they were good hunters, archers, skiers, fishers, sorcerers, magicians and healers.7 We can read about the quality



5 Cf. “Etnicitet eller kulturell identitet lämnar sällan spår i den materiella kulturen” (Werbart 1999, 341) or even stronger: “Det finns ingen relation mellan etnisk grupp och arkeologisk kultur” (Werbart 2002, 29, 102). 6 Both in the classical (from Tacitus, 1 AD) and in mediaeval Latin sources as well as in Old Norse literature, the Sámi were called “finns” (fenni in Tacitus, finnar in Old Icelandic and Old Norwegian written sources). This name for the Sámi has been used until now in some Norwegian dialects. The traditional etymology connects the name finnar with the Germanic root *finþ- (cf. Engl. find). According to this etymology the name finnar designated wanderers, gatherers and hunters (‘finders’). Literature in Svennung 1974, 136–139. 7 Even in Scandinavian medieval and later written monuments in Latin, we can see a lot of information about the Sámi, cf. Historia Norvegiae (12th century, cf. Bäarhielm, Zachrisson 1994), Saxo’s Gesta Danorum (13th century, Saxo 1886) and an especially large amount in Olaus Magnus (16th century), who himself visited the Sámi areas in Sweden. The characteristics of the Sámi in these sources correspond to their characteristics in the Old Norse sources.

69  Sámi and Scandinavians in the Viking Age of Sámi archers in Icelandic sagas and very often these archers are called Finnr. In this connection the name Finnbogi (Finn + bow) is very interesting. The element -bogi (bow) in personal names is possible only with two nations Sámi and Hunns (Hunbogi and Finnbogi). There are no *Danbogi or *Gautbogi or *Gotbogi among Scandinavian personal names. The component -bogi was possible only with people who were known for their archery skills. The bow was a typical Sámi weapon. In a Faroese ballad even Odin appears with a Sámi bow cf. Hár kom maður á völlin fram engin íð hann kendi, eyga hevði hann eitt í heysi finskan boga í hendi. (Hammershaimb, 1851, 11).

The Sámi bow in the hand of Odin points not only to the quality of the bow but also to the capacity of Odin to perform magic. The connection of Odin as magician with the Sámi shamans is clear not only for students of the Scandinavian religion, but even for Loki, who accuses Odin of striking on a drum on Sámsey island as a prophetess, cf. Lokasenna 24: Enn þik síða kóðo Sámseyo í, / ok draptu á vétt sem vo˛lor. The usual interpretation of Samsey as the Danish Island Samsø is hardly correct. In this case we are not dealing with a place name Samsø but with an appellative “a Sámi island”. The Sámi word for selfdesignation sámi / sápmi was known in the Scandinavian tradition. In an Icelandic saga the Sámi sorcerers called themselves semsveinar. The son of the Scandinavian goddess Skadi is called S ming and the Old Icelandic adjective for ‘swarthy, blackish’ sámr is considered to have been borrowed from the Sámi self designation sámi. Sámsey in Lokasenna can be interpreted as a later reinterpretation of an appellative a ‘Sámi island’, which must have originally had the form samey (without -s-, cf. Finney). In this case the original version could look like (“You are known to prophesy on a Sámi island and you stroke on a drum as prophetesses”). But even if the form Sámsey in Lokasenna 24 is original, it is hardly a coincidence that Loki mentioned its name in connection with a typical Sámi procedure – striking the drum during prophesying (Olsen 1960, 19–20).

70  Jurij K. Kusmenko The Sámi capability to prophesy was a very important feature of the Sámi that was attractive to the Scandinavians. The Sámi appear as advisers to prominent Scandinavian personalities, both mythological and historical. Saxo tells us that when Othinus (Odin) asks fortunetellers and priests to give him advice how to avenge the death of his son Baldrus (Baldr), the Finn Rostiophus (Rostiophus Phinnicus) gives him advice how to do it (Saxo 1886, cap. 3, 78). Old Norse sources tell us that even historical Norwegian kings had Sámi as friends and advisors, such as e. g. a Sámi from Hadaland with whom Harald Fairhair flees from his father. Even the first Christian Norwegian King Olaf Tryggvason (d. 1000) visited a Sámi to hear prophecies about his future after his arrival in Norway (Flateyarbók 1, 231). The practice of learning magic and asking for the prophecies of the Sámi was preserved until the 13th century. In two church laws for Eastern Norway (Eidsivaþingslag and Borgarþingslag) one can read that it was forbidden “to go to Sámi” (fara till finna, gera finnfarar), “to believe in Sámi,” (trúa á finna), “to go to Finnmark to ask for a prophesy” (at fara á Finnmerkr at spyrja spá)8, which indicates that in the 13th century this custom was widespread in eastern Norway. The usual adjectives characterizing the Sámi in the Old Icelandic sagas margfróðr, fjo˛lkunnigr, mean not only ‘much knowing’ but also ‘knowing how to perform magic’. To perform magic and to prophesy was not a negative capability before Christianization. On the contrary, it played a very important role in the heathen life of the Scandinavians. The prosaic preface to the Vo˛lundarkviða indicates that even the ability to be a wonderful smith could be connected with Sámi magic power. The preface in prose informs us that the father of the wonderful smith Vo˛lundr was a “Finnish (that is Sámi) king” (finnakonungr). The name of one of Völundr’s brothers was Slagfinnr. All the brothers had a typical Sámi occupation: “they skied and hunted for animals”. The Sámi performance of magic influenced the Scandinavians to a very large degree. In 1935 Strömbäck assumed that “sejd”, the special kind of Nordic shamanism, as it was described in the saga of Erik the Red, had been borrowed from the Sámi (Strömbäck 1935). It is possible that

8 For the texts of these laws see e. g. in Meißner 1942.

71  Sámi and Scandinavians in the Viking Age even the word seid could have been borrowed from Finno-Ugric into the Scandinavian languages.9 The tradition of Sámi magic can be found even in an Icelandic (possibly Norwegian) rune inscription from the 12th century, where the word boattiat in the inscription on a spade shaft boattiat mik inkialtr k rþi was interpreted as an infinitive with imperative meaning from the Sámi verb with the meaning ‘to come’ (North Sámi boahtit). The inscription was interpreted as a spell ‘come back (when stolen or lost)’ (Olsen, Bergsland 1943: 5–7). To use a Sámi magic formula in a spell was quite natural because the Sámi were regarded as the foremost authority in this field. Archaeology shows that representatives of both cultures could marry each other (Zachrisson 1997). Old Norse written sources confirm archaeological findings. The name Halffinnr ‘Sámi by half ’ (formed after the same pattern as a much more known name Halfdan) indicates the Sámi origin of a person, as a rule it was someone who was finnskr at móðurkyni (Sámi after mother) (Pálsson 1999, 31). Old Icelandic sagas tell us that the Sámi women could be wives of legendary and even of historical Swedish and Norwegian kings. A very interesting example is the marriage of the Norwegian King Harald Fairhair to a Sámi woman. Heimskringla of Snorri (13th century) tells us that Harald (d. 933) married a Sámi woman Snæfrid. They had four sons, to whom Harald gave the provinces of Ringariki, Totn and Hadaland (Snorri Sturluson 1941, Ch. 25, 33). There is also a drápa (a verse), which is considered to have been written by Harald after the death of Snæfrid. The motif of marriage of Scandinavian kings to Sámi women can also be found in Gesta Danorum of Saxo Grammaticus. The Danish King Gram declared a war on the Finnish King Sumblus (Sumblus Phinnorum rex), but when he saw his daughter he turned from an enemy into a suitor (Saxo 1886, cap. 1, 18–19). In another story the King of

9 The word seið(r) had many Finno-Ugrian parallels, cf. Sámi sieidi ‘site’ cf. Finn. soida ‘sound, ring’, soitta ‘play on a musical instrument’, Hanty sui, Mansi sei ‘voice’, Ung. zaj ‘noise’; Nenets sjadai ‘wooden Idols’. It is possible that in this case we are even dealing with a much older Finno-Ugrian borrowing into the northern Indo-European languages, cf. Lithuanian saitas ‘sorcery’, saisti (1. Pers. saiču) ‘to read (and interpret) signs’, Welsh hud ( Turkic and Tunguso-Manchurian people > Uralic people (in particular Samoyeds) > Sámi > Scandinavians). In this connection, it is important to stress that the Sámi word bassi (holy) as e. g. in Stoura Bassi Sieidi ‘the great holy god’ (a god of hunt and fishing in one of the Sámi areas) is an Indo-Arian loan word in Finno-Ugric languages (cf. Avest. Baga- ‘luck, fate’, OInd. bhága- ‘luck’).13 The connection of the Scandinavian giant Thjazi with the Sámi god of fishing and water bird hunting Tjatsiolmai becomes even clearer if we remember Thjazi’s daughter Skadi. Skadi has evident features which correspond to the cliché features of the Sámi in Old Icelandic literature. 13 For more detail about the connection between the Scandinavian giant Thjazi and the Southern Sámi god for fishing Tjatsiolmai see Kusmenko 2006.

79  Sámi and Scandinavians in the Viking Age She goes skiing, hunts with a bow and shoots game. In skaldic poetry, she is called ˛ondurdís, ˛ondurgoð ‘ski goddess’. She is a giantess but she belongs to the gods. She was one of Njörd’s wives, but according to the Norwegian skald Eyvind Finnson (!), she did not want to live with Njörd and did better by marrying Odin. She had many children with him. One of them, who was the ancestor of a very well-known person in Norwegian history Hakon Jarl, was called S mingr. Skadi’s traditional occupations and the name of her son with Odin Sæming have given the idea to the known German philologist Karl Müllenhoff that both Skadi and Sæming remind one of the representatives of the original Scandinavian population “Sámi” in Northern mythology (Müllenhoff 1906, 55), cf. also “jettedatteren Skade, som ferdedes på ski, kann godt oprinnelig stamme fra lappenes, ‘finnenes’ saguverden…” (Itkonen 1928, 79). Even the name Sæming is considered to contain the self-designation of Sámi (Sámi / Sápmi). The question is what type of connection is characteristic of the relation between Skadi and the Sámi. Was it a pure coincidence that the features of Skadi coincided with the cliché features of Sámi? Or are we in this case dealing with the personification and mythologization of the northern neighbors of Scandinavians? Such mythologization of the neighbors is a usual thing. The Sámi for instance have also personified and mythologized Scandinavians in the shape of the giant Stallo. Or was Skadi borrowed from the Sámi “tale world”, as Itkonen assumed? We would then have to look for a correspondence for Skadi in the Sámi tradition. Unlike Thjazi, Skadi does not have any formal correspondence among the Sámi gods. But she has a functional correspondence. Among the Sámi akkas ‘female gods’ we find Juxakka, one of Maderakka’s daughters. The name Juxakka can be translated as ‘Bow Woman’. Her attributes are a ski (or a ski pole) and a bow (she is always represented on the shaman drums with a bow and a ski pole). Her function in the 18th century was to be responsible for male children, i. e. for the future hunters. She could even transform a girl into a boy in the mother’s belly. After his birth, she gives the boy to the god of hunting Leibolmai who has the same attributes as Juxakka: a bow and a ski. Leibolmai (‘alder man’) is a male correspondence to Juxakka.

80  Jurij K. Kusmenko Skadi has also a male correspondence: Ull, a god who was called in the skaldic poetry örvaráss ‘arrow-ass’, bogaáss ‘bow-god’, veiðiáss ‘hunt-god’, skíðf rr ‘skier’. Even in this case the connection of the occupations of Ull with the traditional Sámi occupations could not but have caught one’s eye. The German historian Golther stressed that Ull in his armament and his way of life as a hunter with a bow, who goes skiing on snow fields and snow mountains, reminds us of the Finns and the Sámi (Golther 2003 (1895) 13, 312). But it is assumed that the winter features of Ull were not original. In the Scandinavian tradition, Ull is a son of Sif, the wife of Thor, the woman with a golden hair. Though Icelandic mythological tradition does not give us much information about Ull, there are some places in the Elder Edda that show that before the Viking Age he was an important god in Scandinavia, cf. Grm. 42. “Ullar hylli hefr oc allra goða / hverr er tekr fyrstr á funa” ‘Who at first extinguishes the fire has the favor of Ull and of all the gods’ and Atlkv. 30, where an oath must be sworn at the southern sun, at the bourg of Sigtyr (Odin), at the horse of the bed of rest and at the ring of Ull “at sól inni suðrh llo oc at Sigtýs bergi / hölqvi hvílbeðiar ok at hringi Ullar”. Vries indicates Ull as “die helle Seite des Himmelsgottes” whose cult was spread among the Germanic people at the time of the roman emperors (Vries 1957, 159). Ohlmarks dates Ull to even more ancient time. He indicates he was the main god of the sky at the Bronze Age in Middle Sweden and Eastern Norway (Ohlmarks 1975, 181). The OE and OHG personal names with Ull (> Wuld-), cf. OE Wuldwine, OHG Wuldberth indicate his earlier importance in the Germanic world. The runic inscription from Torsbjärg (around 200) owlþuþewaR is especially interesting. This can be interpreted as the servant (i. e. the priest) of Ull (Vries 1957, 158–159). However, the northern features of Ull as a god of winter hunting form a contrast to his original function as a god of light. The most popular etymology connects the name Ull with Goth. wulþus ‘glory, shine’ (Blöndal 1989, 1084). The original meaning of this word is assumed to be ‘light’. Not only etymology but also some places in the Elder Edda (see above) have given rise to the assumption that Ull was originally a god of light and probably one of the most popular gods before the Viking age. Ull was not only connected with Skadi as a god of winter hunting, but this connection can be traced even to

81  Sámi and Scandinavians in the Viking Age the time when he was the god of light. Two circumstances can testify to this. The etymology of the name Skadi, which is connected with Goth. skadus ‘shadow’ and the etymology of the name Ull which is connected with the word for “light”. That the gods of light and shadow formed a pair seems very likely. The second reason for an earlier connection between Skadi and Ull is the geographical distribution of theophoric place names with Skadi (Skadevi, Skädvi, Skädharg etc.), which coincides with the distribution of the theophoric place names with Ull (Ullevi, Ulleraker etc.), (Lindroth 1919, 48; Kraft 2000, 14, 170, 208). The place names with Ull and Skadi are spread in middle Sweden and eastern Norway where hunting was one of the main occupations. Cults of two Sámi gods of winter hunt Juxakka and Leibolmai were spread in the same regions. Thus we have a clear parallel between two Scandinavian gods of winter hunting, a male and a female, Ull and Skadi, and two Sámi gods of winter hunting, a male and a female, Leibolmai and Juxakka. What was the reason for this parallel? We can assume the following development. Before the Viking Age, the North Germanic people had a pair of gods for light (Ull) and shadow (Skadi) or probably for day and night or for sun and moon. It is now impossible to reconstruct the exact meaning of the pair, but it is clear that they formed a pair. The “shadow features” of Skadi have led to her identification with the Sámi goddess of winter hunting Juxakka, cf. the semantic chain shadow > coldness > winter. Through the identification with Juxakka, Skadi was reinterpreted as a goddess of winter hunting (cf. her attributes of a bow and a ski). Accordingly her pendant Ull also obtained the attributes of winter hunting (bow and ski) and also became a god of winter hunting. The transformation of the original god of light Ull into the god of winter hunting is connected not only with the parallelism with Skadi but also with the comparison of the Sámi pair Juxakka – Lieibolmai with the Scandinavian pair Skadi – Ull. In this way Ull received features which were characteristic for Leibolmai. Today it is hardly possible to say how Skadi (a goddess of winter hunting) became a daughter of the giant Thjazi (who originated from the Sámi god of fishing), but it is possible that the functional connection of their Sámi prototypes ( Juxakka and Tjatsiolmai) determined this development.

82  Jurij K. Kusmenko The influence of the Sámi on the Scandinavian sphere of fishing and winter hunting (hunting with skis and bows) must not be a surprise for us. As we have seen above, Sámi had a high authority in these occupations in Scandinavian society. And we must not forget that skiing came to Scandinavians from their northern neighbours (Manker 1971). Language In the period which in linguistic terminology is usually called Common Scandinavian (550–1050) and which in Swedish historical tradition corresponds to the two historical periods Vendel period and the Viking Age, the Scandinavian languages underwent a radical change. At this time they developed several features which distinguish the Scandinavian language from the other Germanic languages but which typologically correspond to the features of the Finno-Ugric languages14. These features include the development of agglutination (suffixed article, suffixed negation, suffixed passive), the loss of the original Germanic prefixes and probably pre-aspiration and nasal assimilation15. These features have always been considered to be the result of an autochthon Scandinavian development. Even Kylstra, who wrote about a typological rapprochement of the Germanic and Finno-Ugric languages did not dare to admit that the typological affinity between the Scandinavian and 14 The increasing morphological affinity of the Scandinavian languages with the Finno-Ugric languages has already been attested to. Kylstra was the first to discover certain important Sámi-Scandinavian parallels. He wrote about “eine deutliche Annäherung des Germanischen an den finnisch-ugrischen Sprachtypus” (Kylstra 1967, 113). He names the following features, which “erinnern… an den Finnisch-Lappischen Typus”: 1. first position of the verb, 2. narrative tense changes, 3. loss of the object, 4. disappearance of prefixes 5. suffixation of the definite article, 6. suffixation of the reflexives, 7. postposition of the possessive pronouns. (ibid, 121). Though some of these features can hardly be connected with Finno-Ugric influence (N 1–3) and others have been named without mentioning possible Finno-Ugric sources of the corresponding Scandinavian developments (N 5–6), this article has made a very important contribution to the study of similar developments in the Sámi and in the Scandinavian languages, though Kylstra himself sees a possible FinnoUgric influence only in the loss of prefixes (Kylstra, 1967, 121). 15 The age of preaspiration can not be established with definiteness because it was never marked in writing.

83  Sámi and Scandinavians in the Viking Age the Sámi languages could be the result of Sámi interference in Common Scandinavian. But if we take into consideration the usual principles of investigation of possible interference features (comparison with related languages, chronology, direction of spreading, degree of incorporation in the language system, typology of interference) we can say that all innovative Scandinavian features mentioned above could be interpreted as the result of Sámi interference in Common Scandinavian. The loss of the original Germanic prefixes was one of the first changes in this direction. The developed system of prefixes was characteristic of the Gothic (cf. prefixes ga-, un, dis-, fair-, twis-) and of the Old West Germanic languages. A very rich system of prefixes is still characteristic of Modern German. Even in English, which has undergone the most radical changes since the Old English period, we find some original Germanic prefixes (cf. become, begin). Common Scandinavian appeared however to be “et praktisk talt prefixløst språk” (Christiansen 1960, 342–343)16. The comparison of the West Germanic languages with Old Norse testifies to this development very clearly; cf. Got. haitan ‘to be called’, gahaitan ‘to promise’, OI heita ‘to be called, to promise’. In some cases, verbs with prefixes in the West and East Germanic languages correspond to verbs with another root, cf. OHG biqueman ‘come up to, get at, become’, OE becuman ‘to reach, to become’ – OI fá ‘to get’. In the position before sonorants, unstressed prefixes have not been completely dropped, they have only lost their vowels as the initial consonant has been incorporated into the root, as in OI granni ‘neighbour’, Got. garazna; OI gnógr ‘enough’, Got. ganohs (vgl. German genug). The loss of the unstressed prefixes can be dated back to the 7th century, cf. the form with the unstressed prefix un- (unnam) in the inscription on the stone from Reistad, 6th century (Krause 1971, 136). If we compare Common Scandinavian with the other Old Ger­ manic languages having unstressed prefixes and Sámi with the other Uralic languages having no prefixes and if we take into consideration that the lack of prefixes (and lack of unstressed initial syllables in general) in Sámi is much older than the Common Scandinavian loss of 16 The new unstressed prefixes of the Modern Scandinavian languages (such as be- and an-) have been borrowed from Middle Low German during the Hansa period.

84  Jurij K. Kusmenko unstressed prefixes, only one direction of borrowing can be assumed, namely from Sámi to Common Scandinavian. The loss of prefixes was originally characteristic of the Scandinavian language of the Sámi but it spread later into the areas without Sámi population. This development in Common Scandinavian corresponded to the main stress pattern of the greatest part of Germanic words which had initial stress. Three other changes in Common Scandinavian (suffixation of -inn, -s(k), -a(t) and -gi / -ki) have also increased the number of iambic words in Common Scandinavian. The traditional hypotheses about the development of the suffixed definite article in Scandinavian languages connects suffixation only with the postposition of the original demonstrative pronoun (maðr + inn góði > maðrinn góði (Grimm 1989 (1898), 447), or maðr inn > maðrinn (Nygaard 1905). The inn-suffixation is traditionally dated to the Viking Age (Noreen 1913). But the postposition of the original demonstrative pronoun is a necessary but in itself insufficient condition for suffixation. It cannot explain the development of suffixation in the Scandinavian languages because in the Old West Germanic languages the postposition of the pronouns was also possible. It seems that there is reason to look for other sources of article suffixation. All the Uralic languages have possessive suffixes which have the same function as possessive pronouns in the Indo-European languages. These possessive suffixes can have not only possessive but also emphatic semantics and semantics that correspond to the semantics of a definite article. On the other hand a definite article very often has possessive semantics, cf. Germ. Ich stecke die Hand in die Tasche vs Engl. I put my hand in my pocket. If we compare languages with possessive suffixes with languages with definite article we can see a clear parallelism between these categories, vgl. Germ. er hat das Bein gebrochen, Sw. han har brutit benet but Finn. hän on murtanut jalkansa, N.Sámi son lea doadjan juolggis. In the Finnish and Sámi sentences the noun for ‘leg’ has a possessive suffix (jalkansa, juolgis), cf. Engl. He has broken his leg. If we compare the semantics of the first cases of inn-suffixation in the Old Scandinavian languages, when the definite article was not yet grammaticalized and the semantics of the suffixed -inn was first of all emphatic and possessive, then the affinity between the Sámi possessive suffixes and the Scandinavian suffixation becomes even clearer.

85  Sámi and Scandinavians in the Viking Age The development of Scandinavian suffixation can thus be reconstructed as follows. Proto-Scandinavian and the start of Common Scandinavian were characterized by free word order in the  noun group. The original Scandinavian demonstrative pronoun (h)inn in postposition was interpreted in the Scandinavian language of the Sámi as a suffix corresponding to the Sámi possessive suffixes which had the same (possessive, determinative and emphatic) semantics. Thus the pronoun (h)it in the Common Scandinavian sentence hann hefr brotit bein (h)it or hann hefr bein (h)it brotit (‘he has broken leg the’) was interpreted as a suffix corresponding to the Sámi possessive suffix -s (cf. son lea doadjan juolggis). It was not the borrowing of a suffix but the reinterpretation of a Scandinavian word as a suffix accordingly to the semantics of the Sámi possessive suffixes. The inn-suffixation, which originally was a characteristic of the Scandinavian language of the Sámi, was later spread to the areas of the original Scandinavian population. Inn-suffixation developed in central Scandinavia, in the main zone of the Sámi-Scandinavian contact, and from there it expanded into the southern Scandinavian area, but suffixation did not reach the southern and western Danish dialects, where the definite article is prepositive.17 The following evidence testifies to the Sámi origin of inn-suffixation: (comparison with related languages) – the presence of the possessive suffixes in all Uralic languages and the absence of article suffixation in the West Germanic languages; (age) – a much younger development of the inn-suffixation in Common Scandinavian compared with the development of the possessive suffixes in the Uralic languages; the spreading of the inn-suffixation from the north to the south (southwest Denmark was not affected by this development). The grammaticalization of the inn-form as a definite article took place later, when the usage of this form with definite semantics became regular. Another case of reinterpretation of Scandinavian postpositive pronouns as suffixes in the Scandinavian language of the Sámi which spread over the whole Scandinavian area is the suffixation of the original reflexive 17 For a detailed study of the rise of the Scandinavian suffixed article and about the article suffixation in other Indo-European languages, see Kusmenko 2001a, 2005).

86  Jurij K. Kusmenko pronouns. The Scandinavian languages, in contrast to the other Germanic languages, have a synthetic passive or middle voice form with the ­suffix -s(k). In the Old Scandinavian languages and in Modern Icelandic, which has best preserved the original status, there is a semantic difference between the middle voice and the analytical passive as in OI opnaðisk ‘opened (by itself)’ – var opnaðr ‘was opened (by someone)’ on the one hand, and a semantic difference between the reflexive and the middle voice (cf. Mod. Icelandic þvo sér (refl.) ‘to wash (oneself)’ þvost (med.) ‘to wash’; baða sig (refl.) ‘to wash (oneself to become clean)’ baðast (med.) ‘to bathe (e. g. in the sun)’(Kress 1982, 198) on the other hand. The grammaticalization of the sk-form indicates that the semantic difference expressed earlier by the same form has acquired morphological significance. The full and the syncopated forms were originally free variants meiða sik – meiðask ‘to get hurt’ (the state that in some cases is preserved in Modern Icelandic setja sig – setjast), but after the grammaticalization of the middle voice, they began to indicate two grammatical categories. The full form began to indicate reflexivity, while the suffixed reduced form was used for middle voice. Scandinavian suffixation can be connected with the reinterpretation of the reduced form of the postpositive Scandinavian reflexive pronoun s(k) < sik, sér as a suffix in accordance with the semantics of the suffix -s in the Sámi languages. The Sámi suffix -s indicates that the action happened by itself (without an agent), cf. Northern Sámi dahpat – dahpasit, (Sw. stänga – stängas), rahpat – rahpasit (Sw. öppna – öppnas), (Nickel, 1990, 228–229). Such a reinterpretation was conditioned not only structurally as in the case of Scandinavian inn-suffixation, but also formally (cf. the phonetic affinity of the Scandinavian -s(k) with the Sámi -s). The Sámi suffix -s is considered to be an autochthon Finno-Ugric suffix (Itkonen 1980, 25) which has correspondences in many Finno-Ugric languages (Estonian and Karelian dialects, Veps, Livonian, Votic, Komi-Permiak – Ariste 1968, 74–75; Aime 1978, 268; Lytkin 1962, 262–266). The Sámi origin of the Common Scandinavian -s(k) suffixation can be testified by:- the structural affinity of the Sámi and Scandinavian voice system (reflexive, middle voice and passive in Common Scandinavian and reflexive, medial and passive verbs in Sámi), the semantics of the -s(k) forms which correspond to the semantics of Sámi medial

87  Sámi and Scandinavians in the Viking Age verbs (especially verbs with the suffix -s), comparison with the related languages (no synthetic voice forms in the West Germanic languages but corresponding forms in other Finno-Ugric languages) and finally the date of the development of the s(k)-form in Common Scandinavian (8th century AD), which is much earlier than the development of the corresponding suffixes in the Finno-Ugric languages. Even in this case we can assume that the appearance of the s(k)-suffixation was initially in the Scandinavian languages of the Sámi in central Scandinavia and later this feature spread to the South.18 Another possible Sámi interference feature in Common Scandina­ vian is the development of suffixed negation, which is dated to the 8th century. The sentence (verbal) negation suffix was -a(t), the sentence constituents negation was -ki / -gi (cf. forms of the Norwegian skalds from the 9th century l trat ‘does not let’, or younger runic munat ‘shall not’ from the 10th century or forms aldrigi ( ne verðrat > verðrat ‘does not become’. We have to understand why this development was possible only in Common Scandinavian. The  suffixation of the  reinforcement elements in Common Scandinavian could have been caused by the reinterpretation of the Common Scandinavian negation construction *ne (negative particle) etR (verb) at (reinforcement) ‘does not eat’ as a construction with a suffix (ne etR a(t) > ne etRa(t)) in correspondence with the Sámi negative construction consisting of a negative auxiliary verb and of a special indefinite (negative) form of the main verb with the original suffix *k > t > 0 (*ejä porek > ej borat > ij bora ‘does not eat’). Thus the Common Scandinavian reinforcements -a ( tt, np > pp, as e. g. in drikka < *drinkan) which occurred in the 7th– 8th centuries (Moberg 1944) and which differed the Scandinavian languages from West Germanic languages could also be connected with a corresponding Sámi development (cf. nG > GG, nD > DD, nB > BB22) which corresponds to the development in some Uralic languages (cf. in Enets – Mikola 2004, 65–66).23 We can see that before and during the Viking Age, Common Scandinavian was strongly influenced by Sámi. The obvious connection between the Scandinavian developments and the Sámi features has always been neglected with the simple reason that the spread of Sámi interference features was not possible due to the low prestige of the Sámi. But we have seen above that relations between the Sámi and the Scandinavians before and during the Viking Age was characterized by mutual respect and acceptance, and did not prevent the spread of Sámi interference features in Common Scandinavian. Conclusion When we look at the relation between the Sámi and Scandinavians in the Viking Age without being prejudiced, we can conclude that the traditional opinion that Sámi cultural influence on the Sandinavians was impossible or minor is false. On the contrary there are many elements in Scandinavian heathen culture and in Scandinavian languages which were borrowed by the Scandinavians from their northern neighbors before and during the Viking Age. We can even suppose that the formation of the Scandinavian culture and of the Common Scandinavian language in the period between the 6th and the 11th centuries was conditioned by Sámi-Scandinavian contacts to a very high degree.

22 Majusculae indicate voiceless lenes plosives. 23 Kylstra, who indicated these parallels between the Sámi and the Scandinavian languages, did not make conclusions about Sámi interference in Scandinavian (Kylstra 1983).

90  Jurij K. Kusmenko Bibliography Aime Kiahrik, 1978: O formirovanii paradigmy vozvratnyh glagolov v vepsskom jazyke. In: Sovetskoje finnougrovedenie, t. 14. 267–276. Ariste, Paul, 1968: A Grammar of the Votic language. Bloomington. Bäärhielm Göran & Zachrisson Inger, 1994: De Finnis, om samerna ur Historia Norvegiae ca 1190 e. Kr. In: Fornvännen 89. 161–162. Baudou, Evert, 1995: Norrlands forntid – ett historisk perspektiv. Stockholm. Baudou, Evert, 2002: Kulturprovinsen Mellannorrland under järnåldern. In: Namn och bygd Bd. 90. 6–36; Blöndal, Magnusson, Ásgeirr, 1989: Íslensk orðsífjabók. Reykjavík. Christiansen, Hallfrid: “De germanske uaksentuerte prefixer i nordisk”. In: Norsk tidsskrift for sprogvidenskap 19 (1960), 340–382. Einarsson, Stefán, 1986(1963): Harp song heroic poetry (Chadwicks). Greek and Germanic Alternative singing. In: Stefán Einarsson, Studies in Germanic Philology. Ed. by Anatoly Liberman. Hamburg. 62–78 Fjellström Phebe, 1985: Samernas samhälle i tradition och nutid. Värnamo. Flat. = Flateyarbók 1860–1868: Unger, Carl Richard: Flateyjarbok: En samling af Norske Konge-Sagaer med indskudte mindre fortaellinger om begivenheder i og udenfor Norge samt Annaler 1. Christiania. Fritzner, Johan, 1877: Lappernes hedenskap og troldomskunst sammenholdt med andre folks, især nordmændenes, tro og overtro. In: Historisk tidskrift. Bd 4. 135–217. Golther, Wolfgang, 2003 (1895): Handbuch der germanischen Mythologie. Rostock. Grimm Jacob: 1989 (1898): Deutsche Grammatik 4. 1. Teil. Besorgt durch Gustav Roethe und Edward Schröder. Hildesheim – Zürich – New York. Haarstad, Kjell, 1992: Sørsamisk historie. Ekspansjon och konflikter i Rørostraktene 1630–1900. Tapir. Hallström, Gustaf: 1921: Gravplatser och offerplatser i ryska Lappmarken. In: Etnolo­ giska studier tillägnade Nils Edvard Hammarstedt. Stockholm. 183–212 Hammershaimb, Venceslaus U., 1851: Sjurdar kv di (Færöiske kvæder, 1 = Nordiske Oldskrifter 12). København 1851 S. 11 ff.). Hansen, Andreas M., 1907: Oldtidens Nordm nd. Ophav og Bos ttning. Kristiania. Hansen, Lars. I., Olsen, Bjørnar, 2004: Samenes historie fram til 1750. Oslo. Hansson, Gunnar Ó., 2001: Remains of a submerged continent. Preaspiration in the languages of Northwest Europe. In: Historical Linguistics 1999. Ed. Laurel J. Brinton. Amsterdam / Philadelphia. 157–173. Helgason, Pétur, 2002: Preaspiration in the Nordic languages. Synchronic and diachronic aspects. Stockholm. Hellquist, Elof, 1993: Svensk etymologisk ordbok. Malmö. Itkonen Erkki, 1980: “Über einige lappische Verbalsuffixe”. In: Suomalais-ugrilaisen seuran aikakauskirja / Journal de la Société Finno-ougrienne 76. 23–41.

91  Sámi and Scandinavians in the Viking Age Itkonen, Toivo Immanuel, 1928: “Fennoskandia-skiens oprindelse”. In: Tromsø museums skrifter; 2. Oslo. 77–87. Itkonen, Toivo Immanuel, 1946: Heidnische Religion und späterer Aberglaube bei den finnischen Lappen. (= Suomalais-Ugrilainen seuran toimituksia; 87). Helsinki. Jahr, Ernst H. 1997. Norway. Kontaktlinguistik. Ed. by Hans Goebl et al. Berlin. 937–948. Kannisto, Artturi, 1919: Zur Geschichte des Vokalismus der ersten Silbe im Wogulischen vom quantitativen Standpunkt. Helsinki. Katz Hartmut, 2003. Studien zu den älteren indoiranischen Lehnwörternn in den uralischen Sprachen. Aus dem Nachlaß herausgegeben von Paul Widmer, Anna Widmer und Gerson Klump. Heidelberg. Koht, Halvdan, 1923: Var “finnane” altid finnar? In: Maal og Minne. 161–175. Koivulehto, Jorma, 2002: Contact with non-Germanic languages 2. Relation to the East. In: Bandle Oskar et al. (Ed.) The Nordic languages. An international handbook of the history of the North Germanic languages. Berlin, New York. 583–594. Kraft, John, 2000: Hednagudar och hövdingadömen i det gamla Skandinavien. Uppsala. Krause, Wolfgang, 1971: Die Sprache der urnordischen Runeninschriften. Heidelberg. Krause, Wolfgang, mit Beiträgen von Herbert Jankuhn, 1966: Die Runeninschriften im älteren Futhark. Göttingen. Kress, Bruno, 1982: Isländische Grammatik. Leipzig. Kusmenko Jurij, 2001a: Die Ursachen der Suffigierung des bestimmten Artikels in den skandinavischen Sprachen. In: Language and Language Behavior. V 4 N 1. St.-Petersburg. 184–206. Kusmenko, Jurij. (2001b): Den nordiska s(k)-formens uppkomst. In: Jurij Kusmenko und Sven Lange (Hrsg.) Nordiska språk – insikter och utsikter. Föredrag från det 5:e studentkollokviet i nordisk språkvetenskap. Berlin. 127–139. Kusmenko, Jurij. (2002): Utvecklingen av suffigerad negation i urnordiskan. In: Jurij Kusmenko und Sven Lange (Hrsg.) Strövtåg i nordisk språkvetensap. Föredrag från det 6:e studentkollokviet i nordisk språkvetenskap. Berlin. 102–118. Kusmenko, Jurij, 2006: Jätten Thjazi och det samiska elementet i nordisk mytologi. In: Andrea Amft & Mikael Svonni (red.) Sápmi Y1K – Livet i samernas bosättningsområde för ett tusen år sedan. Umeå. 11–28. Kusmenko, Jurij, Rießler, Michael. 2000: Traces of Sámi-Scandinavian Contact in Scandinavian Dialects. In: Languages in Contact. Gilbers, Dicky et. al. (Hrsg.). Groningen. 209–224. Kuzmenko Yury, 2005: Saami influence on the development of morphological innovations in Common Scandinavian. In: Finno-Ugric people in the Nordic Countries. Roots V: The roots of people and Languages of Northern Eurasia. Ed. Birger Winsa. Hedenäset. 11–31. Kylstra, Andries Dirk, 1967: Zur Substratforschung. In: Orbis 16 N 1. 101–121. Kylstra, Andries Dirk, 1983: Skandinavisch-Lappische Parallelen. In: Suomalaisugrilaisen seuran toimituksia 185 (Symposium saeculare Societatis Fenno-Ugricae). Helsinki. 159–177.

92  Jurij K. Kusmenko Lindroth, Hjalmar, 1919: En nordisk gudagestalt i ny belysning genom ortsnamnen. In: Antikvarisk tidskrift för Sverige. Bd 20, N 2. 1–76. Lytkin, Vasily, 1962: Komi-permiackiy yazyk. Vvedenie, fonetika, leksika i morfologiya. Kudymkar. Manker, Ernst, 1971: Fennoskandias fornskidor. In: Fornvännen. Bd 66. 77–91. Meißner, Rudolf, 1942: Bruchstücke der Rechtsbücher des Borgarthings und des Eids­ vivathings bearb. von Rudolf Meißner. Weimar. Mikola, Tibor, 2004: Studien zur Geschichte der samojedischen Sprachen. Aus dem Nachlass herausgegeben von Béata Wagner-Nagy. (Studia uralo-altaica 45). Szeged. Moberg, Lennart, 1944: Om de nordiska nasalassimilationerna mp > pp, nt > tt, nk > kk med särskild hänsyn till svenskan. Uppsala. Müllenhoff, Karl, 1906: Deutsche Altertumskunde. Bd 2. Berlin. Mundal, Else, 1996: The perception of the Samis and their religion in Old Norse sources. In: Shamanism and Northern Ecology. Pentikäinen, Juha (Hrsg.). Berlin / New York. 97–116. Mundal, Else, 2004: Kontakt mellom nordisk og samisk kultur reflektert i norrøne mytar og religion. In: Ju. Kusmenko (red.), The Sámi and the Scandinavians. Aspects of 2000 years of contact. Hamburg. 41–53. Nickel, Klaus P., 1990: Samisk grammatikk. Oslo. Noreen Adolf, 1913: Geschichte der nordischen Sprachen. Strassburg. Nygaard, Marius, 1905: Norrøn syntax. Kristiania. Ohlmarks, Åke, 1975: Fornnordisk ordbok. Stockholm. Olrik, Axel, 1905: Nordisk og lappisk gudsdyrkelse. In: Danske studier Bd 2. 39–57. Olsen, Magnus, 1960: Edda- og skaldekvad. Forarbeider til kommentar. II. Lokasenna. Oslo. (Avhandlinger utg. av det Norske Videnskapsakademi i Oslo, II kl. 1960 Nr. 3). Olsen, Magnus, Bergsland, Knut, 1943: Lappisk i en islandsk runeinskrift. (Avhandlin­ ger utg. av det Norske Videnskapsakademi i Oslo, II kl. 1943 Nr. 3). Oslo. Pálsson, Hermann, 1999: The Sami people in Old Norse Literature. In: Nordlit. Arbeidstidsskrift i litteratur. Det humanistiske fakultet. Universitet Tromsø. N 5, vår 1999. 29–53. Qvigstad, Just, K., 1893. Nordische Lehnwörter im Lappischen. (Videnskaps-Selskabs Forhandlingar, 1893:1.) Christiania. Rask, Rasmus, 1932–1933: En Udsigt over de lappiske og finiske Stammers Historie. In: Rasmus Rask: Ausgewählte Abhandlungen 2. Kopenhagen. 285–320. Reuterskiöld, Edgar, 1912: De nordiska lapparnas religion. Stockholm. Rießler Michael, Präaspiration im Nordgermanischen. In: Jurij Kusmenko (Hg.) The Sámi and the Scandinavians. Hamburg 2004. 199–213. Salvesen Helge, 1980: Sørsamene i historisk perspektiv In: Jämten 1981. 127–148. Sammallahti Pekka, 1974: Material from Forest Nenets. Helsinki. Sammallahti, Pekka, 1990: The Sámi Language: Past and Present. In: Arctic languages. Ed. by Dirmid R. F. Collins. Paris. 437–458.

93  Sámi and Scandinavians in the Viking Age Sammallahti, Pekka, 1998: The Saami languages. An Introduction. Karasjok. Sandnes, Jørn, 1973: Om samenes utbredelse mot sør i eldre tid. In: Hiostorisk tidsskrift Bd 52. 113–137. Saxo Grammaticus 1886: Saxoni grammatici gesta danorum. Herausgegeben von Alfred Hodler. Strassburg. Sköld, Tryggve, 1961: Die Kriterien der urnordischen Lehnwörter im Lappischen. Uppsala. Snorri Sturluson, 1941: Noregs konungasögur. Bjarni Aðalbjarnason bjó til prentunar. Reykjavík. Steinsland, Gro, 1986: Giants as Recipients of Cult in the Viking Age? In: Words and Objects. Towards a Dialogue between Archaeology and History of Religion. (ed. by Gro Steinsland). (Instituttet for sammenlignende kulturforskning Serie B: Skrifter 51). Oslo 1986. 212–222 Strömbäck, Dag, 1935: Sejd. Textstudier i nordisk religionshistoria. Stockholm. Svennung, Josef, 1974: Skandinavien bei Plinius und Ptolomaios. Uppsala (Skrifter utg. av k. humanistiska vetenskapssamfundet i Uppsala 45) Vries de, Jan, 1957: Altgermanische Religionsgeschichte. Bd 2. Berlin (2. Aufl.). Vries de, Jan, 1962: Altnordisches etymologisches Wörterbuch. Leiden. Werbart Bozena, 2002: De osynliga identiteterna. Kulturell identitet och arkeologi. Umeå. Werbart, Bozena, 1999: De mänskliga kontakterna i Östersjöområdet. Yngre stenål­ ders fångstsamhällen. In: Göran Burenhult (red.). Arkeologi i Norden 1. Stock­ holm. 336–342. Zachrisson, Inger et al. (red.), 1997: Möten i gränsland: samer och germaner i Mel­ lanskandinavien. Stockholm. Zachrisson, Inger, 2004: Samiskt-nordiska kulturkontakter i mellanskandinavien under 6-700-talen e. Kr. (vendeltid) – utifrån det arkeologiska materialet. In: Ju. Kusmenko (red.), The Sámi and the Scandinavians. Aspects of 2000 years of contact. Hamburg. 9–16. Zachrisson, Inger, 2006: För tusen år sedan – en resa bland sydsamer. In: Andrea Amft & Mikael Svonni (red.) Sápmi Y1K – Livet i samernas bosättningsområde för ett tusen år sedan. Umeå. 75–84.

Þjalfi Anatoly Liberman University of Minnesota

Little is known about Thor’s (Þórr’s) servant Thialfi (Þjalfi or Þjálfi), and this is why the origin of his name, the central subject of this paper, also remains a matter of dispute. An etymologist ignorant of a word’s exact meaning wanders in the dark. The name of an ancient tool (to give a random example) can be explained only when sufficient information exists about the uses to which the tool was put. Likewise, a mythological name (except for such as Freyr ‘lord’ and Þórr ‘thunder’), if treated only as a linguistic sign, is open to all kinds of interpretation. To choose the most persuasive of them, we have to ascertain the character’s place in the system of beliefs. Finn Magnusen 1828: 608–09 compared the name Þjálfi (with á) and the Mod. Icel.1) verb þjálfa ‘to work’, but laid no particular stress on this comparison. Uhland 1868: 33 (first published in 1836) glossed Þjálfi as ‘Arbeiter’, without referring to Magnusen, and is believed to have been the originator of this etymology. Many distinguished scholars repeated it (see, for example, Hermann 1893: 338, Much 1898: 46, Detter 1901: 117, Detter-Heinzel 1903: 221/405, and R. Meyer 1910: 291). Since Uhland’s time several more etymologies of Thialfi’s name have been proposed, and with the exception of one older and one later one, all of them had circulated by 1938 (see a brief survey in Schröder 1938: 214, note 1). Jan de Vries 19772 examines the works of the same authors. Lorenz 1984: 511/14, sec. 2 offers an even shorter survey. The Elder Edda mentions Thialfi only once. In Hárbarðzlióð 37–39, Thor tells Othin (Óðinn) that he had fought “berserks’ brides,” who attacked him with iron cudgels and elto Þjálfa ‘chased away Thialfi’. Berserks’ brides must have been giantesses. No description of this battle has come down to us, but giantesses as Thor’s adversaries figure in Snorri’s Edda, and the theme – Thor battles giantesses and is later mocked for vanquishing women – was popular, as evidenced by several echoes of it in the sagas. Thialfi turns up here as Thor’s companion and, uncharacteristically, flees. Approaching the Viking Age. Proceedings of the International Conference on Old Norse Literature, Mythology, Culture, Social Life and Language. Edited by Ērika Sausverde and Ieva Steponavičiūtė. (Scandinavistica Vilnensis 2). Vilnius University Press, 2009. https://doi.org/10.15388/ScandinavisticaVilnensis.2009.2.6 Copyright © 2019 Authors. Published by Vilnius University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

96  Anatoly Liberman The Lexicon Poeticum (LP) cites two occurrences of the word þjalfi in skaldic poetry. Kormákr used the kennings eyja þjalfi and þangs þjalfi. Both mean ‘sea’ (eyja ‘of islands’, þangs ‘of the sea weed tang’). Guðbrandur Vigfússon (Cl.–V.) was uncertain about how to gloss þjalfi; at present, the kennings are usually translated as “islands’ / sea weed’s encircler, confiner.” The noun þjalmi, the basis of a similar kenning, may be a phonetic variant of þjalfi (see the end of the paper on this word). De Vries 1977 has separate entries for þjalfi and þjalmi. Although he distinguishes between þjalmi ‘rope, noose’ and þjalmi ‘encircler’ (in poetry), he admits that they may belong together. For þjalmi as a prose word Guðbrandur Vigfússon suggested the gloss ‘caltrop’ (that is, ‘snare’) or ‘pitfall’. Þjálmi ‘three loops on a rope’ and þjálmur ‘snare’ have continued into Modern Icelandic. Bugge 1889: 12 also understood þjálmi (with á) as ‘trap, snare’. It follows that Kormákr’s verses do not provide any new information on Þjalfi, for, to interpret the kennings, we have to decide whether they refer to a character called Þjalfi or to the common name homonymous with it. From the skald Eilífr Guðrúnarson’s Þórsdrápa we learn that Thialfi assisted Thor manfully and that neither of them trembled with fear during their encounter with the giant Geirrøðr (strophes 9 and 10). Olrik 1905: 130 believed that the same poet composed Þórsdrápa and Hárbarðzlióð, but no evidence supports this conjecture. According to Snorri, who depended on Þjóðólfr hinn hvinerski (Thiotholf of Hvin), Thialfi did not accompany Thor on his expedition to Geirrøðargarðar but took part in his master’s duel with Hrungnir. Thialfi’s association with Thor must have given rise to numerous versions of their encounters with the giants. He was a good servant, and this is why the line from Hárbarðzlióð comes as a surprise, unless it implies that Thor’s opponents were so dangerous that even the faithful Thialfi fled. Simek 1993 remarks that in Hárbarðzlióð and Þórsdrápa Thialfi may be another god, rather than Thor’s servant. This is probably going too far. Snorri’s version contains a puzzling moment. It is unclear why the giants produced Mo˛kkurkálfi, a clay giant of enormous size. They made him stand near Hrungnir, and Thor’s appearance filled him with great fear. Thialfi attacked, and Mo˛kkurkálfi “fell with little renown.” The situation in which a duel by a minor character precedes or duplicates

97  Þjalfi the main one is known from both heroic literature and myth. Compare Ho˛ttr’s mock fight with a dead dragon after Bo˛ðvarr Bjarki killed it (Hrólfs kraka saga). Even in real fights the moment of triumph resembles an anticlimax. Sigurðr wounded Fafnir from a pit, and Bo˛ðvarr’s encounter with a terrible beast lasted a short time. Only Beowulf and Wiglaf (another pair) struggled in good earnest, and the same holds for the great saga heroes when they defend themselves from mortal enemies. Þjálfi and kálfi ~ kálfr rhyme, but despite the consensus that kálfi means ‘calf ’ (that is, a young bull), Mo˛kkurkálfi need not have been a “calf.” There was a proper name Kálfr, from *Ká-ulfr (De Vries 1977). Therefore, Thialfi may have had a worthy foe to combat rather than a dummy made of clay. Mo˛kkur- in Mo˛kkurkálfi’s name probably meant ‘mist’. The giants (so Snorri) could not find a heart big enough for their creature and put a mare’s heart into him. Unlike Mo˛kkurkálfi, Hrungnir had a sharp-edged three-cornered heart of stone. The duplication of the heart motif has no justification, and it is usually nonfunctional details that provide a clue to the origin of myths. Here we see two of them: a parallel non-heroic duel and a mention of a mare’s heart in Mo˛kkurkálfi’s body. In medieval Scandinavian literature, horses are never denigrated. There is nothing wrong with a mare’s heart, except that it became part of a male (and this is, of course, the whole point). I will venture the hypothesis that in an earlier myth Mo˛kkurkálfi, far from being a huge coward made of clay, was a mighty giant with a stout heart of a stallion (Ká‑ulfr) from the land of the mist, perhaps Hrungnir’s servant. Both had famous hearts, and the battle between Thor and Hrungnir was preceded by a duel between Þjalfi (? Þjálfi) and Kálfi. Thialfi overpowered the servant, and Thor, though not unscarred (a piece of Hrungnir’s hone got stuck in his head and remained there forever) got the better of an almost invincible giant. With time, the story was forgotten and turned into a farce like the gods’ adventure in the kingdom of Útgarðaloki or Ho˛ttr’s attack on a dead beast. E. Meyer’s idea that ‑kálfi means ‘calf of the leg’ (1891: 147), though repeated by von der Leyen 1938: 223, strikes me as fanciful. Our only authority for the beginning of Thialfi’s career is Snorri. According to his tale, Thor set out to visit Útgarðaloki and stopped at a farmer’s house (a typical checkpoint separating people’s habitat

98  Anatoly Liberman from the Other World; see my discussion in Liberman 1994: 185–89). The farmer, his wife, and their two children Thialfi and Ro˛skva invited Thor to a meal. Thor slaughtered his goats and the company partook of them, but Thialfi violated the order not to touch the bones and split the thigh of one of the goats, so that when Thor resuscitated the animals, that goat was lame. He was furious, but, on seeing everybody’s fear, calmed down and did not punish the family. He only took both children with him as servants. All three and Loki show up in Útgarðaloki’s kingdom. Ro˛skva is never heard of again, but Thialfi participates in the contests at Útgarðaloki’s. He claims to be a swift runner, and, although he acquits himself well (he is indeed the swiftest runner on earth), he loses to Hugi (‘Thought’) in all three heats. The lameness of Thor’s goats and the defective handle of Mjo˛llnir belong with Othin’s sacrifice of an eye, Oedipus’s bad foot, Jason’s loss of a sandal, Hephaistos’s lameness, and many other cases of ritual mutilation (they are marks of special destiny) in the myths of the world. In my work on Útgarðaloki, I pointed out an odd detail: the farmer and his wife have no names. In myths everybody and everything (dwarves, giants, swords, kettles, etc.) has a name. Consequently, when we are told that, for instance, Baldr’s nameless horse was buried with him or that Thor dined with an anonymous farmer, we may suspect that Baldr did not have a horse and that such a farmer never existed. Rydberg 1886: 642 (and seemingly, no one else) commented on the farmer’s anonymity, but he had a penchant for imaginative cross-references and did not realize that we are dealing with multiple versions of a fluid tale rather than fragments of a fixed text. It is said in Hymisqviða, stanza 7, that when Thor decided to visit Hymir, the giant Egill took care of the goats. Rydberg concluded that the maiming of the goats happened during Thor’s absence and that Thialfi was Egill’s son. Snorri, in his opinion, did not know the name of Thialfi’s father (he says that in the Younger Edda the name was forgotten). This reconstruction is groundless. Thor would not have sought a servant among the giants’ sons. But in one respect Rydberg was right. Snorri’s account does contain a flaw: Thialfi could not be “some farmer’s” son either. An additional source used in unraveling the Thialfi myth is the Old Swedish Guta saga. It begins with a legendary description of Gotland,

99  Þjalfi an enchanted (eluist, that is, elvist / elfist) island that sank into the ocean at night and came to the surface during the day. The power of elves was broken and the sinking stopped when a man called Thielvar (Þieluar, that is, Þielvar) came to Gotland with fire (quam…eldi). Nothing is known about his wife; however, he had a son Hafþi who married Huitastierna (that is, Hvitastierna); she appears in the legend from nowhere. On their first night, she had a prophetic dream, and so on. The inhabitants of Gotland are the offspring of that first couple. Since only one personage in Scandinavian myths is called Thialfi, researchers identify him with Thielvar. I will look at this tradition in some detail, but its shaky foundation should be pointed out at once. Compare Läffler’s reservations (1908: 171) and Finnur Jónsson’s cautious remark: “Some people think that Thialfi is identical with Thielvar” (1913: 59). Peel 1999: xviii is also noncommittal. Since Thialfi was Thor’s servant, and Thor, as his name indicates, was originally a thunder god, an attempt has been made to treat Thialfi as lightning. It seems to have been E. Meyer’s idea (1891: 204; 1903: 277), but Much developed it like no one else (1898: 46 = p. 55 of the book edition). The episode in Guta saga allowed him to compare Thialfi / Thielvar with Prometheus. However, the comparison is weak. Thielvar did not steal fire from the gods or bring it to people (Gotland was uninhabited). His action had a different purpose: to take possession of land, one had to carry fire over it (hence the phrases koma eldi and fara með eldi um landnám). This is what happened to Gotland: once Thielvar performed the ritual, the elves lost their power over the island and the area became his property (cf. Rydberg 1887: 103, who held this opinion). Detter 1893: 116, in a review of E. Meyer 1891, asked: “Why should Thor’s servant Thialfi be lightning?” The reason is clear but wrong. Some mythologists treated Loki as an ancient fire demon or fire god, stressed the proximity of Thialfi’s and Loki’s roles (both often accompany Thor), and thus justified their treatment of Thialfi / Thielvar, the alleged Scandinavian counterpart of Prometheus. One of them was Schröder 1924: 117. Later we will see that Loki and Thialfi have nothing in common, but Schröder’s views deserve more than a passing mention, for he gave the problem of Thialfi his serious consideration. Regrettably,

100  Anatoly Liberman when he returned to it (1938: 212–15; 220; 221, note 2; 222), he offered an eccentric hypothesis. He repeated his old idea that Thielvar was a fire demon but now added that he was the son and husband of Gotland’s mother goddess. A long passage follows to the effect that such a situation is common in mythology. Huitastierna, Thielvar’s daughter-in-law, turned out to be his wife, identical with Tacitus’s Nerthus. Her name (‘White Star’), so Schröder, may mean ‘die blondgelbe (Acker)fläche’, approximately ‘blond-yellow (plowland)’ but, most probably, means ‘a cow with a white star on the forehead’; she was, he proposes, a cow goddess (this is an old conjecture). In other cases Schröder also reconstructs theriomorphic deities. For example, from Thor’s association with goats he concludes that at one time Thor was a god in the shape of a goat. “Beyond any doubt, the Gotland creation tale is based on the idea that the god of fire and fertility forms a union with the earth goddess. This is the first couple, and all people are their offspring” (215). Now that, according to Schröder, not a single dark corner has remained with respect to Thialfi’s nature, the origin of his name, as he believes, also becomes clear. Presumably, Gmc. *þelfēn goes back to PIE *telp‑ or *telbh‑, whose root is (s)tel‑ ‘drip, urinate’, as in Gk σταλάσσω ‘drop, drip’, G stallen ‘urinate’ (said about horses), and others. More cognates can be found in WP 642–46, especially on p. 646. The connection between G stallen and Gk σταλάσσω is questionable, but the Greek verb and Gmc *þelf (whatever its meaning) may be related. However, the similarity between ejaculation of semen expected from a fertility god (or urination) and dripping ~ dropping is distant, to put it mildly. Þjalfi is a weak noun (an n‑stem), while Þielvar is not. Schröder could not account for the difference and referred to the possibility of an ancient alternation in the suffix (n ~ r). Schröder was not fully confident of his etymology and dismissed the question with the statement that the origin of mythic names plays a subsidiary role in understanding the origin of myths. As to the myth, he thought he had found support for his reconstruction in the fact that Thialfi was Thor’s servant. Another long passage is devoted to Indra and Vishnu and Indra as an ithyphallic god (‘god with an erect phallus’) and to other gods, mainly Greek, having the same characteristics. The excursus led him to the conclusion that when a great god is accompanied by a

101  Þjalfi small servant (Schröder emphasizes the servant’s size), the god embodies the force of procreation, whereas the servant (as far as his origin is concerned) is the god’s worshipped anthropomorphic phallus. By way of afterthought (221, note 2), Schröder mentions the Hittite vegetation god Telepinu ~ Telebinu, whose name looks like a perfect congener of Gmc *þelf‑, but adds that more research is needed to make this comparison valid and promises to return to the ties between Germanic and the languages of Asia Minor. In his later works, Thialfi does not seem to have surfaced again. The first to represent Thialfi as small was probably Olrik (1905). He objected to the understanding of Thialfi as a worker, because Thialfi was a short, weak, even though nimble companion of the thunder god (138). Olrik mentioned two circumstances relevant to his interpretation: in myths, strong gods are seldom smart (he could have referred to the usual folklore juxtaposition of brawn versus brain, especially prominent in animal tales) and need small resourceful servants; besides, loud thunder peals are usually preceded by weaker, more distant ones (138–39). Considering that in the extant corpus of Scandinavian myths even Thor has nothing to do with thunder (only his name means ‘thunder’, and his hammer, especially if Mjo˛llnir is related to Russ. molniia ‘lightning’, resembles a thunder god’s weapon) and that, as we will see later, according to Olrik’s untenable proposition, Loki, rather than Thialfi, was Thor’s original servant, his hypothesis holds out no promise. Nor did he connect Thialfi’s name with thunder. He thought that Þjalfi was perhaps a variant of Þialfar and decomposed the latter name into þial, the Old Swedish form with breaking corresponding to OI þel ‘ground; strength’, and far‑, as in the verb fara. The whole yielded ‘precipitous runner’ (138). In Scandinavian myths, unlike what one encounters in later romances and the eddic verses based on them, in which monsters with 900 heads threaten the protagonist, the giants, dwarves, and gods are anthropomorphic, and their stature depends on their status: the giants are dangerous (and this eventually made them look big in people’s eyes), the dwarves are the gods’ servants (this factor contributed to their becoming diminutive), while Thor is a giant slayer (and hence towering over everyone else). Other than that, they interact as physically equal

102  Anatoly Liberman beings. The dwarf Alvíss woos Thor’s daughter, Freya sleeps with four dwarves, dwarves overpower a giant, Thor experiences no discomfort while staying in a farmer’s house, and so forth (cf. Liberman 2008: 47–49; the beginning of the entry dwarf). Pictorial representations of Thor as big and Thialfi as small reflect the dichotomy master / servant, not tall / short. To boost the idea of Thor’s little servant, Schröder (1938: 219, end of note 4 from the previous page) glossed Lytir, the name of an obscure Swedish divinity (he wrote Lýtir), as ‘der Kleine; Däumling’. Neither Lytir nor Lýtir (a less probable variant) is related to OI lítill ‘little’ (see Liberman 2008: 144, the end of the entry lad). Von der Leyen 1938: 223 called Þialfi one of the most delicate (zierlichsten!) gods. The passage in his book devoted to Thialfi is unfortunate. He followed E. Meyer and considered Mo˛kkurkálfi to be a misty calf of the leg and Thialfi a quick sunray piercing the mist. Not a single source calls Thialfi small, weak, nimble, or delicate. Those are fancies. The epithets applied to him are sjálflopti ‘self-soaring’ (Eilífr) and fóthvatari ‘swifter of foot’ (than anyone else; Snorri). Comparative mythology reaches its lowest point when it allows itself to be carried away by wide-ranging convergences. The topic at hand tends to be lost in a display of erudition. We only know that Thielvar brought fire to Gotland, broke the spell the elves had laid on it, and had a son Hafþi by an unknown mother, who in turn married a woman called Huitastierna. That couple populated the entire island. In Guta saga, Thor does not appear. In Old Icelandic poetry (skaldic and eddic) and prose (Snorri), Thialfi is Thor’s servant. Where are the fire demon, the mother goddess, the ithyphallic Thor, his anthropomorphic phallus, and a son sleeping with his mother? Some of them occur in other religions, and that is where they should stay. I prefer to treat with equal disbelief Lemke’s dream-symbolic explanation of Thielevar’s activities (1986: 13–16). As already mentioned, Olrik believed that Thialfi ousted Loki from the  place of Thor’s servant. In an Estonian tale (he noted) it is the trickster who is inseparable from the thunder god: he gets him into trouble (cf. the stealth of Iðunn’s apples) and rescues him (cf. Þrymsqviða). Also, since, according to Olrik’s theory, a strong god of limited intelligence needs a smarter helper, Loki appears to be qualified

103  Þjalfi for the resourceful servant’s role, whereas Thialfi does not (the ruse that secured Thor’s victory over Hrungnir was suggested by the gods, not by him). In Þrymsqviða, Loki flies to Thrymr’s kingdom with Thor, and in the adventure at Útgarðaloki’s Thor, uncharacteristically, has two companions: Loki and Thialfi, in addition to Ro˛skva (Olrik 1905: 138, 140–46). Olrik was consistent and called Thialfi’s participation in the Hrungnir myth a late detail (130). This reconstruction disregards several moments. To begin with, no single role fits Loki, and he is too important to be a mere companion of another god (see Liberman 1994 for a full discussion of Loki’s character and his development from a chthonian deity). It may be true that someone like Thor needs a smart servant to offset his simple-minded brutality, but he also needs someone who will help him win battles (a Wiglaf at the side of a Beowulf), and Loki despite his participation in the Ragnaro˛k, an all-out confrontation between order and chaos, is not a fighter. Secondly, the eddic gods regularly travel in groups. Alongside Odin and Thor, we sometimes see obscure figures like Hœnir and Lóðurr (for instance, in Vo˛lospá 18). Occasionally Loki bears them company. Olrik’s idea appeals to those who treat Loki as a primordial fire demon (compare what has been said above about Thialfi / Thielvar ‘lightning’). Few scholars were ready to substitute Loki for Thialfi in the latter’s capacity as Thor’s servant. (See Celander 1911: 90–92; De Vries 1937, II: 45, and Ström 1956: 51; Lorenz 1984: 510/11, sec. 3 mentions the controversy but does not discuss it. Neither does De Vries, who only registers his disagreement with Olrik.) Philippson 1953: 48–49 suggested that Loki as Thor’s companion supplanted the colorless Thialfi, though Þórsdrápa gives Thialfi his due. In reality, two independent lines – Thor / Loki and Thor / Thialfi – must have crossed at the earliest time. Thor had many companions but only one known servant (Thialfi). Ro˛skva may have been another, but her mythology is lost. Lindow 2001: 286 writes: “Following Georges Dumézil, many observers, especially those who, like Dumézil, approach the material from the Indo-European side, see here [in Thjálfi’s failed attempt to lift Hrungnir’s lifeless leg off Thor] a reflection of warrior initiation: Under the tutelage of an elder warrior, the initiant slays a made monster. I find the theory attractive even though there is nothing in Snorri’s

104  Anatoly Liberman text to indicate that Thjálfi’s status changes after the encounter with Mökkurkálfi, which we would expect in an initiatory context. Made monsters turn up in all sorts of cultures, not always in initiatory contexts (e. g. golems).” Indeed, the farce, mentioned above in connection with Ho˛ttr, looks like a parody of initiation, while the episode involving Mo˛kkurkálfi does not. The loss of the Ro˛skva myth deprived us of valuable material, for some conclusions regarding Thialfi could have been drawn from the character of his sister. Ro˛skva does not participate in the games at Útgarðaloki’s, and in Old Icelandic poetry her name appears only twice as part of kennings, in which it functions as a trivial synonym for “woman” (LP). Etymologists have interpreted Ro˛skva as a cognate of OI ro˛skr ‘brave’ (this connection is beyond dispute) and possibly of OI roskinn ‘ripe, mature’ (related to Go. gawrisqan ‘bear fruit’). Ro˛skr may have begun with *h‑ (though an h‑less variant has also been recorded), while roskinn began with *w‑; for this reason, they cannot be related, unless we resort to the formulation that they were “variants” of the same root. Þórsdrápa (stanza 21) has Vro˛sku, not improbably, a deliberate archaicization of the name on analogy with roskinn rather than a decisive argument for the Ro˛skva‑gawrisqan connection. In the older scholarly literature, in which Ro˛skva was most often understood as a cognate of gawrisqan, her name led to recognizing an ancient goddess of growth and fertility. Since a thundergod controls clouds and rain, protecting crops also falls within his jurisdiction. But nothing testifies to the role some nineteenth-century researchers ascribed to Ro˛skva. Her name, despite Eilífr’s Vro˛sku, is, more likely, related only to ro˛skr. The distant origin of ro˛skr is of no importance in the present context, for even in Eilífr’s days roskin had no v‑ and ro˛skr had no h‑ and speakers could not distinguish between the two roots. OE r scan ‘quiver, flash’ and the development of ro˛skr (rask means ‘quick’ in all the modern continental Scandinavian languages) show that the semantic kernel of *raskur was ‘impetuous, energetic’ (this is the meaning of Mod. Icel. röskur); Engl. rash (from Scandinavian) and G rasch have made the same way as rask in Norwegian, Swedish, and Danish. It seems more natural to assume that Ro˛skva meant ‘an energetic, impetuous, brisk one’ (such

105  Þjalfi is also the opinion of nearly all later scholars) and that she could run as well as her brother Thialfi. (Is this the reason she has nothing to do at Útgarðaloki’s?) One of the oldest interpretations of Thialfi’s function and name appears in Cl.–V. Here is the relevant part of the entry: “Þjalfi… the name of the servant and follower of Thor. Edda; and also as a pr[oper] name; the word prop[erly] means delver, digger, Germ. delber, delben, = to delve, dig; the names Þjálfi and Röskva… indicate that Thor was the friend of farmers and the god of agriculture.” For completeness’ sake, I will also reproduce the entry about Ro˛skva: “Röskva,… rhymed Vröskva, Þ[órs] d[rápa]; the name of the maiden follower of Thor; she is a personification of the fields of harvest” (the next entry is röskvask ‘to grow up, to ripen’). Þjalfi and delve (from OE delfan) cannot be related, for an Old English cognate of Þjalfi would also have had initial þ, as Detter 1893: 117 observed, but he made his point so quickly that to notice it is hard. Guðbrandur Vigfússon, the author of the Þjálfi‑delve etymology, found an enthusiastic supporter in Rydberg, who spun one of his imaginative reconstructions around it. As we remember, according to Rydberg, Thialfi was the son of the giant Egill and the splitting of the goat’s thigh allegedly happened in his house. Rydberg recalls a fornaldrasaga in which Gróa finds a little boy in a fl ðarmál and brings him up (a fl ðarmál is the space between low and high water), cites several other episodes of the same type, and concludes that the name Þjalfi suggests a similar idea. He identifies Þjalfi, the name, with þjalfi ‘sea encircler’ and calls Thialfi a discoverer of lands and a circumnavigator of islands. Unexpectedly, he interpreted the verse in Hárbarðzlióð, in which “berserks’ brides” are said to have chased away Thialfi, as meaning that Thialfi conquered giants, to make the land inhabitable. This, naturally, brought him to Guta saga. Thialfi emerged as a delver, ready to work the ground with a spade (Rydberg 1886: 708–10). Thialfi did not delve anything, but as will be shown, one of Rydberg’s ideas can be salvaged. I am aware of five more comments on Thjalfi. Loewenthal, the author of numerous mechanical etymologies, connected Þjalfi with Russ. dialectal tolpega ‘lout’ (stress on the second syllable; the pejorative meaning comes from the suffix). Since, according to Olrik, the concept

106  Anatoly Liberman of the Germanic fire demon (Feuergeist) developed from “he who brings fire,” in Loewenthal’s view (1921: 261–62), Þielvar may be understood as ‘someone who forces his way [with fire]’. The root of tolpega is tolp‑ (cf. Russ. tolpa ‘crowd’, stress also on the second syllable; the older meaning of the root must have been ‘to have enough room’). No tie can be detected between tolpega and Þjalfi, and all the important questions, such as the relationship between Thialfi and Thielvar, remain unanswered in this etymology. Gordon 1927: 178, note to line 116, says that Þjálfi (sic) is probably identical with Þieluar, “who took fire to Gotland and so disenchanted it. The name means ‘one who seizes and holds’, and is etymologically identical with þjálmi ‘receptacle’, ‘noose’” (the same in the second edition by A. R. Taylor, 1957, and in the 1962 reprint from corrected sheets, p. 199). The equation Þjálfi = þjálmi is old, but the gloss ‘one who seizes and holds’ that Gordon gives must have justified, in his opinion, Thielvar’s “seizing and holding” of Gotland. No one seems to have shared Gordon’s idea. Gutenbrunner 1936: 159–60 suggested that Þjalfi is an abbreviated form of Þjelvar (from *Þelba‑harja‑). He assumed their identity with þjalfi ~ þjalmi ‘snare; fetter’ and referred to the custom of some warriors among the Chatti of wearing rings; hence his gloss Þielvar ‘fettered fighter’ (Fesselkämpfer). It remains unclear whether the common name þjalfi is also an abbreviation of some longer word, for if it is not, then the argument falls to the ground. Besides, neither Thialfi nor Thielvar was fettered or wore a ring as a mark of belonging to a religious union, and neither deserves the name of a Weihkrieger (? ‘an initiated warrior’). Therefore, this interpretation can also be dismissed as unrealistic. Mogk traced Þjalfi to *Þjalfr (a strong form alternating with a weak one), which allowed him to combine the Icelandic name with Þielvar, and etymologized it as *Þewa‑alfaR. In his additions to the commentary on the Elder Edda, Gering 1927, II: xviii mentioned this etymology without discussion, and Sturtevant 1952: 1147 found it the best there was. (Did Mogk suggest *Þewa‑alfar to Gering in a letter? In the first, 1891, edition of Paul’s Grundriß, p. 1093, Mogk wrote “Þjálfi, d. h. der Gräber,” that is, “delver,” “probably lightning going into the ground” [der in die Erde fahrende Blitz], and referred to “a popular Germanic

107  Þjalfi myth,” according to which fire is lightning coming out of the ground. He glossed Mo˛kkurkálfi as ‘a heavy cloud’ and Ro˛skva as ‘a quick one’. The text in the 1908 edition, p. 358, is the same. Mogk’s 1898 book is an offprint of his chapter from the Grundriß: see p. 132 on Thialfi. In Hoops’s Reallexikon [Mogk 1918‑19: 323], Þjálfi again appears with á, but no gloss follows his name. Ro˛skva is ‘a quick one’, and Mo˛kkurkálfi is not mentioned. Sturtevant gives no reference either.) Sturtevant saw no difficulties in the  derivation *þewa‑alfaR  > *þē‑alfr > *þjalfr but realized that Þjalfi was not an elf (“Þjalfi and Ro˛skva were of peasant origin”). His explanation runs as follows: “…from the standpoint of the relation between Thor and Þjalfi as ‘master’ and ‘servant’, it does not seem inappropriate to call the ‘servant’ an ‘elf ’, inasmuch as the elves, like the dwarfs, were subservient to the ­gods–note that Freyr became king of the elves and resided in Alf‑heimr…, an abode which the gods had already bestowed upon him. The name Þewa‑alfaR might then be equated with the name Þór‑alfr… as an elf who was in the service of Thor. Mogk’s derivation offers no more serious difficulties than do the other derivations discussed by Gering (I: 250).” Although this derivation may be better than the others, the ancient metaphor (a servant called an elf) does not inspire confidence. Our sources say little about the elves in Scandinavian myths. Their status has been examined many times, most recently by Árman Jakobsson 2006: 229–38. No revelations came to light, and, however broadly we may wish to interpret the material, the statement that the elves, along with the dwarves, were subservient to the gods would be hard to confirm. Hall’s 2007 book should also be consulted. His theme is Anglo-Saxon and later beliefs, but he cast his net broadly. Finally, Gust Johansson 1969 discussed all the names in Guta saga and said that Þielvar might perhaps be understood as Tjäll‑vard ‘hut guardian’ (18; the Modern Swedish form of the name is Tjälvar). The element ‑var is irreconcilable with vard (for what happened to ‑d?), and in what sense was Thielvar, let alone Thialfi, the keeper of a house? Surveys of scholarship can be found in three works: Gering 1923–27, I: 250; Schröder 1938: 215, note 1, and De Vries 1977 (Þjalfi). De Vries missed Olrik and mistakenly presented E. Meyer as the originator of the Þjalfi-delve etymology (the reference should have been to Cl.‑V.);

108  Anatoly Liberman besides, the comparison Þjalfi – Mod. Icel. þjálfa goes back to Magnusen, not Uhland. With respect to þjálfa, Schröder made the same mistake. The other etymological dictionaries of Icelandic offer nothing new. Holthausen 1948 and Alexander Jóhannesson 1956: 447 compare Þjalfi and þjálfa and derive both from the root *telp‑. ÁBM follows their example but mentions some of the older conjectures. It should be added that the style of some etymological works deserves little praise. Hypotheses are often rejected with one word (“improbable,” “erroneous,” and the like). Schröder called Olrik’s derivation a Verlegenheitserklärung, that is, ‘an explanation offered when there is nothing to say’, ‘an explanation born of despair, a face saving device to avoid embarrassment’. Gutenbrunner suggested his gloss ‘fettered warrior’ without examining the opinions of his predecessors, as though they did not merit even a passing negative remark. By contrast, one’s own view is usually promoted most forcefully and all difficulties are treated as relatively insignificant. I also have a proposal, but, before launching it, I will give a short summing up. Old Icelandic poets and Snorri remembered Thialfi only as Thor’s servant. Someone often dupes and helps the thunder god (it may be the same person), but, in dealing with Thialfi, I will disregard the evidence of comparative religion, rely on the meager information at our disposal, and proceed on the assumption that Thialfi had always been Thor’s servant and did not supplant Loki or anyone else in that role. Like Freyr, who had a male and a female servant (Byggvir and Beyla), Thor had Thialfi and Ro˛skva in his service (brother and sister), but no tales of Ro˛skva seem to have circulated in the North even in the 10th century. While accompanying Thor, Thialfi killed Mo˛kkurkálfi, a giant. Thialfi enjoyed enough popularity in Sweden, for his name to be used among mortals. Þielvar may be a variant of the same name. The likelihood of their identity is great. Yet the two mythic characters share no obvious common features, unless we identify Thialfi with lightning (which is inadmissible). Thielvar did not serve Thor (or any other god) and did not fight giants. He was the culture hero of Gotland: he took possession of the island, cleansed it of evil spirits, married Huitastierna (a woman of unknown antecedents), and became the progenitor of Gotland’s population. Huitastierna is indeed a typical cow name, but the distance

109  Þjalfi between the figure in Guta saga and a cow goddess is too great for us to span. (Likewise, no path leads from Thor, as he appears in the extant Scandinavian myths, to an ancient goat god.) Some time after I formulated my proposal concerning the origin of Thialfi’s name, I noticed a reference to Jacob Grimm in E. Meyer 1891: 204 (“hardly the same as Thjolf, Donarulf ”) and learned that Grimm had not missed Thialfi. (In Deutsche Mythologie, he is not mentioned.) It turned out that I had partly reinvented his etymology (a common case in etymological studies). In a footnote to an article on the names of thunder (Grimm 1865: 409, note, continued on p. 410; first published in 1853), he asked whether Thialfi could not be understood as Donnerwulf ‘thunder wolf ’, for he assisted Thor in carrying his cudgel (Grimm took kyll ‘bag’ for a cognate of G Keule ‘club, cudgel, a weapon of a thunder god’) and his sister also made a lot of noise (he connected Ro˛skva with OI raska ‘displace’, that is, ‘cause disorder’). Grimm, quite naturally, was aware of Magnusen’s works but may not have consulted his entry Þjálfi, for he found þjálf ‘work’ not in Magnusen’s Lexicon but in Bjørn Haldorsen’s 1814. He thought of Go. þei o ‘thunder’ and ulfr ‘wolf ’. It is a brilliant etymology: perfect with respect to meaning and flawless from a phonetic point of view (cf. Go. lei an ‘lend’ and OI ljá < liá). *Þeih‑ulfr would have yielded Þjálfr. This idea must have occurred to Grimm too late to be incorporated into the 1854 edition of his Deutsche Mythologie. My starting point was such Old Icelandic names as Hrólfr, Bjólfr, Jólfr, Þjóðólfr (see them in De Vries 1977, under ulfr), and So˛kkólfr (Noreen 1923: sec. 130). All of them have ‑ólfr from ulfr. (This, however, does not hold for Gylfi, in which lf belong to the root; its likeliest cognate is OI gjalfr ‘sea’. See Olrik 1910: 12, Finnur Jónsson 1934–35: 294, Björn Sigfússon 1933: 131, and Sturtevant 1940–41: 223–24.) I assume that the original form of the name was *Þjalfr, with Þjalfi considered to be more familiar and more appropriate for a servant. Strong and weak forms of the same name often existed side by side: cf. Yngvar and Yngvi (an extreme case). So many ingenious etymologies of Þjalfr have led nowhere because everybody, except Jacob Grimm, tried to explain Þjal‑fi instead of Þja‑lfi. However, Grimm’s *Þeih‑ulfr was not the best choice, because Go. þei o has no cognates in Germanic; consequently,

110  Anatoly Liberman such a Scandinavian word for “thunder” (to the best of our knowledge) did not exist. Here I think Mogk guessed well: the first element was *þewa‑, as in OI þjá ‘serve’ (< *þewan). Thus, Þjalfi came into being as *Þewa‑ulfr (‘serve’ + ‘wolf ’), an ideal name for a servant. The only hitch is the vowel length. It seems that *Þewaulfr should have become Þjálfr. This form exists but is believed to be secondary, due to vowel lengthening before certain consonant groups, one of which was lf‑ (Noreen 1923: sec. 124.3 and 237.2). Sturtevant, famous for his attention to phonetic detail, reconstructed, as noted, the string *þewa‑alfaR > *þē‑alfr > *þjalfr (he also assumed the primacy of the strong form) and said nothing about why Þjalfr had short a. *Þewaulfr must have yielded *Þiáulfr ~ *Þjáulfr, with contraction in hiatus producing a long vowel. Is it possible that Þjálfi was an original form and that the skalds used the variant Þjalfi to fit the meter and that á in this name did not emerge as ú did in úlfr ‘wolf ’ (< ulfr), thanks to lengthening? Considering the fact that lf functioned as a lengthening group, we should expect some vacillation. In my opinion, the development of a in Þjalfi, though problematic, need not derail the protoform *Þewaulfr. Once upon a time *Þewa‑ulfr encountered *Mo˛kkur‑ulfr and vanquished him. He deserved to enter into Thor’s service. It becomes clear why the young man’s greatest virtue was “great speed.” A servant is first and foremost a messenger and is expected to be everywhere in no time. His sister was also quick, rather than a fertility goddess. Compare the etymology of OI þr ll ‘slave, servant’. Usually the forms OE þr gan and Go. þragjan ‘run’ are given as the most secure cognates of þr ll (a good semantic parallel in Greek, first suggested by Brugmann, can be found in Feist 1939, þragjan). Russ. sluga ‘servant’ (stress on the second syllable) has no accepted etymology, but, according to one of the proposals, it may be related to the words with the root *sel‑ ‘move, flow’ (see the end of the entry sluga in the Russian translation of Vasmer’s dictionary: III, 676). The question about the relation between Thialfi and Thielvar defies a definitive answer. With the fire lightning motif discredited, only their names remain a connecting element between them. Yet a certain detail may rescue their affinity in myth and legend. One was allowed to take as much land as one could carry fire over or plow from sunrise to

111  Þjalfi sunset (see Olrik’s discussion of this ritual in various cultures in Olrik 1910: 4–8). If Thielvar had not run all the way around Gotland in one day, after sunset it would again have sunk into the ocean. To perform such a task, he had to be a very swift runner, for the island is large. Perhaps this is the reason Thialfi was chosen as Gotland’s culture hero, but the link is admittedly weak. The identity of Thialfi and Thielvar should be neither denied nor made too much of; see also what is said below on OI þjalfi. Mod. Icel. þjálfa ‘to work hard; train’ was recorded only in the 17th century, and þjálf ‘training’ was derived from it (ÁBM). A verb with such a meaning would probably have surfaced in Old Icelandic if it had existed in it. Most likely, both þjálfa and þjálf are late formations, so that it would be wrong to trace Þjalfi to them. If Þjalfi reflects the mythic character’s nature (as *Þewaulfr is supposed to do), the common name þjalfi ‘encirler, confiner’ is not related to it. Its etymology constitutes a problem of its own, but the coexistence of the homonyms Þjalfi ‘Thor’s servant, the best runner in the world’ and þjalfi ‘encircler’ inevitably affected the meaning of both or at least of the proper name. If Þjalfi came to mean ‘encircler’ (in addition to ‘servant; runner’), the kenning eyja þjalfi would have merged with eyja Þjalfi ‘an “encircler,” or “Thjalfi” of islands’. This is where Rydberg’s ‘circumnavigator of islands’ may come in. In the kenning, an encircler, confiner of islands was the sea, but, if applied to a human being, it would have fit the activities of someone like Thielvar, who “encircled” Gotland from within, rather than from without. This is one more argument for identifying Thialfi and Thielvar. The whole, of necessity, remains guesswork. The skalds used þjalmi ‘rope; snare’ as a doublet (variant) of þjalfi. The correspondence OI þjalfi ~ OIr tailm ‘snare’ cannot be fortuitous. Bugge 1889: 12, ever on the lookout for the Celtic influence on Old Icelandic, cited OIr tailm ~ teilm (genitive telma) and Welsh telm (the same meaning) and explained the name Þjálfi (sic) as a borrowing from Old Irish. He did not distinguish between Þjálfi and þjálfi and equated Þjálfi with Þjálmi. His idea of borrowing has nothing to recommend it; note only that he gives a in all those forms length (Mogk and Neckel‑Kuhn, as we have seen, also wrote Þjálfi). Noreen 1923: sec. 237.2 was not sure whether the variation þjalfi ~ þjálmi went

112  Anatoly Liberman back to phonetic reasons: the Celtic forms lend þjalmi an independent existence and make the picture unclear. OI þjalfi, whatever its origin and whatever its relation with OIr tailm, seems to be a different word from Þjalfi ~ þjálfi, but, once their paths crossed, they could not help beginning to interact. Notes 1. The following abbreviations are used in the text of this paper: Engl. – English, G – German, Gk – classical Greek, Gmc – Germanic, Go. – Gothic, Mod. Icel. – Modern Icelandic, OE – Old English, OI – Old Icelandic, OIr – Old Irish, PIE – Proto-Indo-European, Russ. – Russian. 2. When a word occurs in a dictionary, page numbers are not given.

Bibliography ÁBM = Ásgeir Blöndal Magnússon, Íslensk orðsifjabók. Orðabók Háskólans, 1989. Alexander Jóhannesson. See Jóhannesson, Alexander. Árman Jakobsson. See Jakobsson, Árman. Björn Sigfússon. See Sigfússon, Björn. Bugge, Sophus. 1889. “Om Versene i Kormaks saga.” Aarbøger for nordisk oldkyndighed og historie II/4, 1–88. Celander, Hilding. 1911. Lokes mytiska ursprung. Språkvetenskapliga sällskapets i Uppsala förhandlingar, Sept. 1906 – Dec. 1909. Uppsala: Akademiska Boktryc­ keriet Edv. Berling. Cl.–V. = An Icelandic–English Dictionary initiated by Richard Cleasby, subsequently revised, enlarged and completed by Gudbrand Vigfusson…. Oxford: Clarendon Press, 1874. 2nd ed., with a supplement, by William A. Craigie, 1957. De Vries, Jan. 1937. Altgermanische Religionsgeschichte. Vol. 2. [Pauls] Grundriß der germanischen Philologie 12/II. Berlin and Leipzig: Walter de Gruyter. ———. 1977. Altnordisches etymologisches Wörterbuch. 3rd ed. Leiden: E. J. Brill. Detter, Ferdinand. 1901. [Rev. of] Meyer, Elard Hugo. 1891. Zeitschrift für deutsches Altertum. Anzeiger 19, 113–19. Detter–Heinzel = Ferdinand Detter and Richard Heinzel (eds.). 1903. S mundar Edda mit einem Anhang. Vol. 2: Anmerkungen. Leipzig: G. Wigand. Feist, Sigmund. 1939. Vergleichendes Wörterbuch der gotischen Sprache. 3rd ed. Leiden: E. J. Brill. Finnur Jónsson. See Jónsson, Finnur. Gering, Hugo. 1923–27. Kommentar zu den Liedern der Edda. Posthumous edition by B. Sijmons. Halle (Saale): Buchhandlung des Waisenhauses.

113  Þjalfi Gordon, E. V. 1927. An Introduction to Old Norse. Oxford: Clarendon Press. 2nd ed., by A. R. Taylor, 1957 (repr. 1962, etc.). Grimm, Jacob. 1865. “Über den Namen des Donners.” Kleinere Schriften. Vol. 2: Abhandlungen zur Mythologie und Sittenkunde. Berlin: Ferd. Dümmler, 402–38. (First published in 1853.) Gutenbrunner, Siegfried. 1936. “Zur Gutasaga.” Zeitschrift für deutsches Altertum 73, 159–63. Haldorsen, Bjørn (= Björn Halldórsson). 1814. Lexicon islandico-latino-danicum…. Havniæ: Apud J. H. Schubothum. Hall, Alaric. 2007. Elves in Anglo–Saxon England: Matters of Belief, Health, Gender and Identity. Woolbridge, Suffolk UK; Rochester, NY: Boydell Press. Halldórsson, Björn. See Haldorsen, Bjørn. Hermann, Paul. 1903. Nordische Mythologie in gemeinverständlicher Darstellung. Leipzig: Wilhelm Engelmann. Holthausen, Ferdinand. 1948. Vergleichendes und etymologisches Wörterbuch des Altwestnordischen. Göttingen: Vandenhoeck & Ruprecht. Jakobsson, Árman. 2006. “The Extreme Emotional Life of Vo˛lundr the Elf.” Scan­ dinavian Studies 78, 227–54. Jóhannesson, Alexander. 1956. Isländisches etymologisches Wörterbuch. Bern and München: Francke. Johansson, Gust. 1969. Guta saga. Lekmannafunderingar. Göteborg: Nordiska Boktryckeri. Jónsson, Finnur. 1913. Goðafr ði norðmanna og íslendinga eftir heimildum. Reykjavík: Gutenberg. ———1934–35. “Þulur: søkonge og jættenavneremserne.” Acta Philologica Scandi­ na­vica, 289–308. Läffler, L. Fr. 1908. “Till 700-årsminnet of slaget vid Lena (31 Januari 1208). 3. Ett stadgande i Gutasaga, som ytterst föranledts of slaget vid Lena. En laghistorisk undersökning.” Fornvännen 3, 137–177. Lemke, Uwe. 1986. Gotaland, Insel der Götterschiffe. 2nd ed. Stuttgart: Urachhaus Johannes M. Mayer. Leyen, Friedrich von der. 1938. Die Götter der Germanen. München: C. D. Beck. Liberman, Anatoly. 1994. “Snorri and Saxo on Útgarðaloki, with Notes on Loki Laufeyjarson’s Character, Career, and Name.” In his Word Heath. Wortheide. Orðheiði: Essays on Germanic Literature and Usage (1972–1992). Roma: Il Calamo, 176–234. (First published in 1992.) ———. 2008. An Analytic Dictionary of English Etymology: An Introduction. (With the assistance of J. Lawrence Mitchell.) Minneapolis, London: University of Minnesota Press. Lindow, John. 2001. Handbook of Norse Mythology. Santa Barbara, CA; Denver, CO; Oxford, England: ABC-CLIO. Loewenthal, John. 1921. “Religionswissenschaftliche parerga zur germanischen altertumskunde.” Beiträge zur Geschichte der germanischen Sprache und Literatur 42, 239–65.

114  Anatoly Liberman Lorenz, Gottfried. 1984. Gylfaginning / Snorri Sturluson: Texte, Übersetzung, Kommen­ tar. Texte zur Forschung 48. Darmstadt: Wissenschaftliche Buchgesellschaft. LP = Sveinbjörn Egilsson, Lexicon Poeticum antiquae linguae septentrionalis. Ordbog over det norsk-islandske skjaldesprog. 2nd ed. by Finnur Jónsson. København: S. L. Møller, 1913–16; reprinted København: Atlas Bogtryk, 1966. Magnusen, Finn. 1828. Prisc Veterum Borealium Mythologice Lexicon. Havniæ: Typis Jani Hostrup Schulz. Meyer, Elard Hugo. 1891. Germanische Mythologie. Lehrbücher der germanischen Philologie 1. Berlin: Mayer & Müller. ———. 1903. Mythologie der Germanen gemeinschaft dargestellt. Strassburg: Karl J. Trübner. Meyer, Richard 1910. Altgermanische Religionsgeschichte. Leipzig: Quelle & Meyer. Mogk, Eugen. 1891. “Mythologie.” [Pauls] Grundriss der germanischen Philologie. Vol. 1. Strassburg: Karl J. Trübner, 932–1138. ———. 1898. Germanische Mythologie. 2nd ed. (Sonderdruck aus der zweiten Auflage in Pauls Grundriss der germanischen Philologie.) Straßburg: Karl J. Trübner. ———. 1900. “Mythologie.” [Pauls] Grundriss der germanischen Philologie. Vol. 3. Straßburg: Karl J. Trübner, 230–406. ———. 1918–19. “Thor.” [Hoopses] Reallexikon der Germanischen Altertumskunde. Vol. 4. Straßburg: Karl J. Trübner, 322–24. Much, R. 1898. “Der germanische Himmelsgott.” Abhandlungen zur germanischen Philologie. Festgabe für Richard Heinzel…. Halle a. S.: Max Niemeyer, 189–278. (Published as a separate book by the same publisher in the same year with a name index.) Neckel-Kuhn = Gustav Neckel, ed., Edda. Die Lieder des Codex Regius nebst verwandten Denkmälern. Heidelberg: Carl Winter, 1914. 3rd and 4th eds. by Hans Kuhn, 1962–68. Noreen, Adolf. 1923. Altnordische Grammatik I…. 4th ed. Halle an der Saale: Max Niemeyer. (Reprinted as the 5th ed. Tübingen: Max Niemeyer, 1970.) Olrik, Axel. 1905. “Torenguden og hans dreng.” Danske Studier, 129–46. ———. 1910. “Gefjon.” Danske Studier, 1–31. Peel, Christine, ed. 1999. Guta Saga: The History of the Gotlanders. Viking Society for Northern Research. Text Series, vol. 12. University College London: Viking Society for Northern Research. Philipsson, Ernst A. 1953. Die Genealogie der Götter in Germanischer Religion, Mythologie und Theologie. Illinois Studies in Lan­guage and Literature 37/3. Urbana: The University of Illinois Press. Rydberg, Viktor. 1886. Undersökningar i germansk mythologi. Part 1. Stockholm. Albert Bonnier. ———. 1887. Fädernas gudasaga. Stockholm: Albert Bonnier. (Reprinted Uddevala: Niloe, 1984.) Schröder, Franz Rolf. 1938. Germanentum und Hellenismus. Untersuchungen zur germanischen Religionsgeschichte. Germanische Bibliothek II/17. Heidelberg: C. Winter.

115  Þjalfi ———. 1924. “Germanische Urmythen.” Archiv für Religionswissenschaft 35, 201–36. Sigfússon, Björn. 1934–35. “Names of Sea–Kings (Heiti S konunga).” Modern Philo­ logy 32, 125–42. Simek, Rudolf. 1993. Dictionary of Northern Mythology. Translated by Angela Hall. Cambridge, England; Rochester, NY: D. S. Brewer. Ström, Folke. 1956. Loki. Ein mythologisches Problem. Göteborgs Universitets Årsskrift 62/8. Göteborg: Elander. Sturtevant, Albert Morey. 1940–41. “Some Etymologies of Old Norse Poetic Words.” Scandinavian Studies 16, 220–25. ———. 1952. “Etymological Comments upon Certain Old Norse Proper Names in the Eddas.” Publications of the Modern Language Association 67, 1145–62. Uhland, Ludwig. 1868. Der Mythus von Thor nach nordischen Quellen. In his Schrif­ ten zur Geschichte der Dichtung und Sage. Vol. 6: Sagenforschungen. Stuttgart: J. G. Cotta. (First published in 1836.) Vasmer, Max. 1950–58. Russisches etymologisches Wörterbuch. Heidelberg: Carl Winter. Russian edition: M. Fasmer, Etimologicheskii slovar’ russkogo iazyka. Translated and enlarged by O. N. Trubachev…. Moscow: Progress, 1964–73. WP = Alois Walde, Vergleichendes Wörterbuch der indogermanischen Sprachen. Posthu­mous edition by Julius Pokorny. Berlin: Walter de Gruyter, 1927–32. (Reprinted 1973.)

Norwegian Modal Verbs and Attitudinal Modality Ugnius Mikučionis Vilnius University

Fregna ok segja skal fróðra hverr, sá er vill heitinn horskr… Hávamál

Den som gjerne vil gjelde for klok, skal utveksle nytt med andre… Håvamål

In this article, I present a semantic model of modality, where the starting point is the assumption that modality is a semantic category which deals with people’s attitude towards the trustworthiness of propositions and / or the desirability of states of affairs. Terminology In the following, I will simply use the term ‘attitude’ rather than repeating ‘attitude towards the trustworthiness of propositions and / or the desirability of states of affairs’, since these are the only areas I treat as modal. In my understanding, modality is not concerned with people’s attitude towards other properties, such as good / bad, clever / stupid, easy / difficult, big / small, important / insignificant, cheap / expensive, and so on. Only trustworthy / untrustworthy (of propositions) and desirable / undesirable (of states of affairs) are truly modal. As a general term, I will use OK-ness, covering both trustworthiness and desirability. As technical terms, I will use epistemic attitude when talking about the evaluation of the trustworthiness of a proposition, and non-epistemic attitude when talking about the evaluation of the desirability of a state of affairs. Approaching the Viking Age. Proceedings of the International Conference on Old Norse Literature, Mythology, Culture, Social Life and Language. Edited by Ērika Sausverde and Ieva Steponavičiūtė. (Scandinavistica Vilnensis 2). Vilnius University Press, 2009. https://doi.org/10.15388/ScandinavisticaVilnensis.2009.2.7 Copyright © 2019 Authors. Published by Vilnius University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

118  Ugnius Mikučionis The word epistemic derives from Ancient Greek έπιστήμη meaning ‘knowledge, science’. However, the epistemic kind of modality deals with what is believed to be true rather than what is known to be true. This follows from my definition of modality as an attitudinal category, and also harmonizes with most other current definitions of ‘epistemic’. In this respect, then, the term is somewhat unfortunate. However, it is so conventional in the linguistic literature that I do not see any point in trying to replace it with a new term. The speaker may refer to her own or to someone else’s attitude, and I will use the term participant which also has become conventional in modern linguistic literature on modality. As several authors have pointed out (e. g. van der Auwera & Plungian 1998: 83; Andersson 2007: 13f.), it is to be preferred to the alternative term, agent (used by, e. g., Bybee et al. 1994), since the person referred to does not necessarily have the agent role in an actual utterance – it may also be the patient, benefactive or have some other role. I shall treat attitude as a notion with two values, neutral and positive. These can be combined with negation to form negative attitude. Negative attitude is thus seen as a composite category. Negation may be treated as a separate feature or factor, which may be added to modal expressions in utterances, so that this combination expresses the participant’s negative attitude. This is obviously correct with respect to the use of modal verbs, such as must, shall, can or will, all of which can be combined with the negative particle not. Although there exist lexical items, such as scarcely, hardly, prohibited which represent the participant’s negative attitude, I choose to analyze the negative attitude as a complex value, resulting from a combination of (non-neutral) attitude and negation. Therefore, I will only operate with two types or degrees of attitude – positive (non-neutral) and neutral, which partly correspond to the traditional terms necessity and possibility, used in most literature on modality. However, there are some important differences between what is called neutral attitude and possibility, and to an even greater extent between positive attitude and necessity, as I will try to demonstrate in the following paragraphs. Neutral attitude means that the speaker has no objections to accept a proposition as correct or a state of affairs as worth to occur. However,

119  Norwegian Modal Verbs and Attitudinal Modality she may equally accept that the same proposition may turn out to be incorrect, or the same state of affairs may turn out not to be worth to occur. In either case, no problems (no conflict) will arise for the speaker with respect to her beliefs or expectations. Positive attitude means that the speaker is willing to accept a proposition as correct or a state of affairs as worth to occur. If the proposition turns out to be incorrect, or the state of affairs turns out not to be worth to occur, a conflict arises between the speaker’s beliefs and / or expectations and the reality, i. e. there is a problem. However, the speaker may indicate in the utterance that she admits that other attitudes are possible. To put it in other words, the speaker may signal that other participants may have different attitudes than her own, but this does not mean that the speaker is unsure about her own attitude (if this were the case, one would have to do with neutral attitude, cf. above). As technical terms, I will use simple and complex attitude to distinguish between cases where the speaker in her utterance expresses only her own positive (non-neutral) attitude without approving of any alternative attitudes, and cases where she expresses her own positive attitude at the same time as she signals in the same utterance that other participants may have a different attitude towards the status of the attitude target, i. e., the proposition or the state of affairs in question. The distinction between simple and complex attitude is only relevant in connection to non-neutral attitude. Neutral attitude is automatically simple since the speaker does not – and cannot – invite anyone to a discussion or to negotiations about the trustworthiness of a proposition or about the desirability of a state of affairs. Such discussion or negotiations are only possible when the speaker has a non-neutral attitude and is willing to listen to alternative attitude(s). In practice, the complex attitude may be perceived as a lower degree of commitment on the part of the speaker toward the attitude target, as compared with the simple attitude which sounds more categorical and uncompromising. Thus, the complex attitude can easily be interpreted as containing a certain element of doubt or non-assuredness and consequently as representing lesser confidence from the side of the speaker. However, the speaker does not actually need to be unsure

120  Ugnius Mikučionis about her own attitude in order to be able / willing to allow the other participant(s) to express their (alternative) attitude.1 As to terminology, traditional terms in the literature on modality are necessity and possibility. However, there are reasons to avoid using them in the description of my model of modality. It would not be logical to distinguish between simple (non-negotiable) vs. complex (negotiable) necessity, as the term necessity refers to something absolute and undisputable. The distinction between something the speaker agrees to dispute and something she does not, is an essential part of my model of the semantics of modality. Therefore I stick to the use of the terms neutral and positive attitude in preference to possibility and necessity. Actually, many authors have operated with terms like strong vs. weak obligation and tentative vs. confident conclusion, which point in the direction that modality is treated as a gradable domain or even as a continuum (van der Auwera & Plungian 1998: 82; van der Auwera et al. 2005: 251–252). But since it is not logical to talk about strong (confident) necessity / possibility as opposed to weak (tentative) necessity / possibility, the terminology gets unnecessarily complicated. To my mind, the cleaner the terminology is, the more adequate the analysis one can achieve. I choose therefore to abandon the traditional distinction between necessity and possibility, in favour of talking about different types of attitude and complexity, which are represented in figure 1 below.

1 As an alternative set of terms for the description of this distinction between the different types of non-neutral attitudes, I have considered negotiable vs. non-negotiable attitude. These terms also represent the idea of the speaker’s willingness (or unwillingness) to accept alternative attitudes. There is however some risk that the reader will misinterpret these terms as necessarily invoking some actual negotiations between the speaker and the other participant(s) in a concrete communication situation. The idea here is actually that the speaker may signal her readiness or willingness to accept alternative attitudes in the very same utterance where she expresses her own attitude. Nothing is said about whether any actual negotiations between two or several participants will ever take place. Therefore, I have finally chosen to use the terms simple vs. complex attitude throughout the article; of course, the adjective “simple” has here nothing to do with “simple-minded”, “simplistic” or “naïve (attitude)”.

121  Norwegian Modal Verbs and Attitudinal Modality

attitude

neutral

complexity

simple

positive complex

Figure 1. Types of attitude and complexity.

As already mentioned above, the term attitude is used in connection with the speaker’s evaluation of the trustworthiness of propositions (epistemic attitude) and her evaluation of the desirability of a state of affairs to occur (non-epistemic attitude). When there is no need to specify whether epistemic or non-epistemic attitude is meant, I use the term OK-ness which covers both trustworthiness and desirability. This leads to the preliminary representation of modal domains given in figure 2. Non-epistemic attitudes (evaluation of desirability of states of affairs)

Epistemic attitudes (evaluation of trustworthiness of propositions)

Complex positive attitude (in the speaker’s view it is OK only if the state of affairs occurs, but the speaker signals in the same utterance that there is room for alternative attitudes)

Complex positive attitude (in the speaker’s view it is OK only if the proposition turns out to be correct, but the speaker signals in the same utterance that there is room for alternative attitudes)

Simple positive attitude Simple positive attitude (only OK if the state of affairs occurs) (only OK if the proposition turns out to be correct) Neutral attitude (OK if the state of affairs occurs, but also OK if it does not)

Neutral attitude (OK if the proposition turns out to be correct, but also OK if it does not)

Figure 2. A preliminary representation of modal domains.

122  Ugnius Mikučionis In the following sections I will discuss the different types of attitude in the epistemic and non-epistemic domain in greater detail and provide examples that may serve as empirical evidence that my model of modality is not only based on theoretical considerations, but also represents linguistic reality well. Finally, I would like to mention that I have chosen to use the pronoun ‘she’ when referring to the speaker, and ‘he’ when referring to the hearer or other participant(s) in a communication situation. Non-epistemic modality Strictly speaking, modality refers to the attitude in both epistemic and non-epistemic modality. In the case of non-epistemic modality, the attitude is pragmatically connected with expectations about the participants’ actions and therefore with certain speech acts. This is, in principle, a secondary effect. These speech acts are determined not only by modality itself, but also by the communication situation, which in its turn is primarily determined by the (number of) participants involved. Nonepistemic modality differs from epistemic modality in that it is connected with reactions and / or actions, besides describing the participants’ attitude. Epistemic modality is only connected with the participant’s attitude towards the OK-ness of a proposition or a state of affairs. In the traditional literature on non-epistemic (specifically deontic) modality, one usually speaks about permission and different types of so-called mands (commands, demands, encouragements, requests, entreaties), which are called non-epistemic (deontic) possibility and non-epistemic (deontic) necessity, respectively. In my view, permission, encouragement or command are not different (sub)types of nonepistemic modality, but rather different types of speech acts, the use of which depend both on the speaker’s (or some other person’s) attitude and on the communication situation. The neutral attitude in connection with non-epistemic modality (non-epistemic attitude) carries the meaning that, in the participant’s view, there are no obstacles for the state of affairs to occur – it is OK that the state of affairs occurs, but it is also OK if it does not occur. The reason for why it is OK that a state of affairs occurs does not need to be expressed in an utterance. Thus, the reason is actually not relevant

123  Norwegian Modal Verbs and Attitudinal Modality for the identification of the attitude as neutral. This is not to say that the reason is of no relevance for the choice of lexemes in concrete utterances, as we will see in the examples below. The non-neutral attitude in connection with non-epistemic modality means that, in the participant’s view, matters are OK only if the state of affairs occurs. If the state of affairs fails to occur, there is a conflict between the participant’s expectations or interests and reality. An utterance may, of course, contain certain information about the obstacles for a state of affairs to occur. Likewise, one can state that no obstacles are present in a given situation. The speaker may for example indicate that there is no prohibition (either by the speaker herself or by law) or that there are no physical, material obstacles for the state of affairs to occur. An utterance may also contain information about reasons for why it is important to ensure that a state of affairs does not fail to occur. Such information may be explicitly expressed by lexical means, but may also be indicated by the speaker’s choice of modal verb. A typical example from Modern Norwegian is the use of the modal verb får in utterances expressing permission. Du får gå nå (= you may go now, you are permitted to go now) differs from Du kan gå nå (= you can go now) in terms of explicitness regarding the obstacles. The latter utterance simply indicates that there are no obstacles for the participant to leave (without indicating what kind of obstacles could prevent him from being able to do so), while the former indicates that there is no prohibition (= the obstacle) to leave. By using the verb får, the speaker thus grants her own permission or refers to someone else’s permission for the participant to leave. Another typical example is the use of the modal verb skal in utterances expressing command, with 2nd person subject. Du skal gå nå (= you are obliged to go now, you are commanded to go now) differs from Du må gå nå (= you must go now) in that the latter utterance more neutrally indicates that the participant is forced to leave (one does not indicate what kind of circumstances force him to do so), while the former indicates that it is someone’s will (= the obstacle) which requires that the participant leaves. A similar difference may be observed between utterances containing the modal verb bør as compared to utterances containing the modal

124  Ugnius Mikučionis verbs skal or må. Du bør gå nå (= you ought to go now) means that, in the speaker’s view, it is in the other participant’s own interests that he goes now. By such an utterance, the speaker sends a signal that other people may have different attitudes towards whether the matters really only are OK provided that the person in question goes now. In practice, such an utterance may be interpreted as expressing a lower degree of confidence on the part of the speaker regarding what her own attitude actually is. Both Du skal gå nå (= you are obliged to go now, you are commanded to go now) and Du må gå nå (= you must go now) disallow any other points of view towards the OK-ness of the matters and consequently can be interpreted as expressing a higher degree of confidence on behalf of the speaker. As already mentioned, in the traditional literature on modality one usually speaks about different (sub)types of non-epistemic modality, depending on the nature of the obstacles. The most established notions are participant-external vs. participant-internal modality, and dynamic, deontic and boulomaic (boulethic) modality. In my view all of these notions refer to different communication situations, rather than constituting different types of modality. By communication situation, I mean first of all the number of participants involved in a conversation. This determines the nature of the speech (monologue or dialogue), and allows for variation regarding the source of attitude, that is, the person whose attitude is being reported. The speaker does not necessarily report her own attitude. By separating modality, which deals with the attitude, from information about the kinds of obstacles or reasons behind the attitude, we are able to avoid the major problem with the traditional interpretation of modality, namely, the difficulty of proving that all the different subtypes are actually parts of the same linguistic category. In other words, we avoid the difficulties by formulating a definition of modality which is equally well suited for all the subtypes of modality, and at the same time excludes other categories, such as tense or aspect. The neutral attitude in connection with non-epistemic modality, as has already been pointed out, may be paraphrased as no obstacle for a state of affairs to occur, or it is OK if a state of affairs occurs (but also OK if it does not).

125  Norwegian Modal Verbs and Attitudinal Modality The positive (non-neutral) attitude in connection with non-epistemic modality may be paraphrased as the matters are OK if – and only if – a state of affairs occurs, but the speaker may signal that it is “allowed” for other people to have different attitudes towards the same state of affairs. Such a “democratic” attitude may be expressed explicitly by lexical means or by the choice of modal auxiliary in a concrete utterance. Let us now take a look at different types of attitude in connection to non-epistemic modality. If the speaker holds that there are no obstacles for a state of affairs to occur, she may say a sentence like one of the following. 1. Neutral attitude in connection to non-epistemic modality a) Du kan reise til Paris. ‘You can go to Paris.’ b) Du får reise til Paris. ‘You may (are allowed to) go to Paris.’ c) Du må gjerne reise til Paris. ‘It is fine with me if you go to Paris.’ d) Bare reis til Paris, du! ‘Just go to Paris!’

All these utterances may be paraphrased as It is OK if you go to Paris (but also OK if you don’t). The utterance in (1a) is the most unspecified one in the sense that it simply states the absence of obstacles for the state of affairs to occur, saying nothing about the nature of the obstacles. (1b) expresses permission, that is, absence of prohibition, which constitutes information about the obstacle. (1c) and (1d) sound most natural in situations where the speaker reacts to the other participant’s attitude, that is, where the other participant has shown his own non-neutral attitude towards the state of affairs (in this case, willingness to go to Paris) and the speaker is now giving her approval. If the speaker considers that matters are OK only if the state of affairs occurs, she may use an utterance like one of these.

126  Ugnius Mikučionis





2. Positive (non-neutral) attitude in connection to non-epistemic modality simple attitude a) Du må reise til Paris. ‘You must go to Paris.’ b) Du skal reise til Paris. ‘You shall go to Paris.’ c) Du vil reise til Paris, altså. ‘So, you will (= want, wish) to go to Paris.’ d) Reis til Paris! ‘Go to Paris!’ complex attitude e) Du bør / burde / skulle reise til Paris. ‘You ought to / should go to Paris.’ = It is desirable with respect to your own interests that you go to Paris (it is not OK for yourself unless you go to Paris).

All these utterances may be paraphrased as It is OK if, and only if, you go to Paris = It is not OK unless you go to Paris. The utterance in (2a) is the most unspecified one in the sense that it only shows the speaker’s attitude towards the state of affairs, namely, that the participant must leave for Paris in order for matters to be (become / stay) OK. Nothing is said about the reasons for why it has to be so. The utterance in (2b) means that someone has planned the participant’s journey to Paris. This utterance thus contains information about the nature of the obstacle. The utterance in (2c) means that the reason for why it is necessary to go to Paris, is the participant’s own will. The utterance in (2d) is a command and will typically be used in situations where the speaker has authority to decide what is desirable and what is not. The utterance in (2e) differs from the utterances in (2a–d) in that it signals that the speaker dissociates herself from being the only licit source of norm: the decision to go to Paris or not rests with the other participant himself. Thus, the speaker expresses her point of view quite unambiguously, but at the same time (and in the same utterance) she

127  Norwegian Modal Verbs and Attitudinal Modality indicates that the other participant(s) can have different attitudes. In the traditional literature on modality such utterances are said to express weak obligation, but in my view it is more precise to analyze them as representing the speaker’s complex attitude towards the state of affairs. To put it in other words, the speaker does not express that, in her view, it is less necessary for the other participant to go Paris by choosing the auxiliary bør (ought to) instead of må (must) or skal (shall). What the speaker does say, is that she is not expressing the only possible attitude towards the necessity of the journey to Paris. Thus, the choice of a modal verb in a concrete utterance may reflect considerations which are primarily related to the identity of the source of attitude and the complexity of the attitude (simple or complex). The consequence of the pragmatic interpretation of these considerations is assigning the utterances in (2a–e) different degrees of OK-ness of the state of affairs. The source of modality need not be expressed in an utterance. If we only have an utterance like Du må reise til Paris, and no additional information, we can say that the attitude is non-neutral and simple, but the source of attitude is unspecified. Optative is another example where the source of attitude remains unspecified. 3. Unspecified source of modality, optative a) Leve kongen! ‘Long live the king!’ b) Må kongen leve lenge! ‘May the king live long!’

The attitude represented by such utterances is non-neutral and simple. In my model, then, non-epistemic modality partly overlaps with what has traditionally been called boulomaic (boulethic) modality, deontic modality and dynamic modality, or, in terms elaborated by van der Auwera and Plungian (1998), participant-external and participantinternal modality. The non-epistemic modality also covers wishes and fears, which are treated as partly deontic and partly epistemic by Palmer (2001: 13).

128  Ugnius Mikučionis Epistemic modality As mentioned, epistemic modality is pure attitude, in my view. Unlike non-epistemic modality, it does not require any reaction and does not involve any actions from the side of the participants of the communication situation. The speaker expresses (her own or someone else’s) attitude towards the trustworthiness (likelihood) of a proposition. As in the case with non-epistemic modality, one can speak of the source of attitude here, as well. The speaker is identical with the source of modality when she expresses her own attitude, and not identical with the source of modality when she refers to someone else’s attitude.2 The source of modality can also remain unspecified. The neutral attitude in connection with epistemic modality simply means that the participant has no reason to believe that the proposition is wrong or false; there is no obstacle to accept the proposition as potentially true. It does not say anything about whether the participant expects the proposition to be true or not; it says only that it may well be true (but may also be false). The  most typical means of expressing the  neutral attitude in English are utterances containing modal verbs can and may, while in Norwegian it is utterances with the modal verb kan. Such utterances can usually be paraphrased by It is possible that…, It is possibly the case that…, or, to include the notion of obstacle, by There is no obstacle to assuming that the following proposition is true (although it may also be false). 4. Neutral attitude in connection with epistemic modality Han kan ha reist til Paris. ‘He can have left for Paris.’ = It is possible that he has left for Paris. = There is no obstacle to assuming that he has left for Paris.

The positive (non-neutral) attitude in connection with epistemic modality means that the participant has essential willingness (not reason – although willingness may be invoked by some reason, it may also remain unmotivated) to believe that the proposition is correct; or, to

2 Reporting someone else’s words is usually considered a kind of evidentiality.

129  Norwegian Modal Verbs and Attitudinal Modality employ the notion of obstacle, the participant will encounter obstacles to accepting any other conclusion than the proposition being correct. The nature of the obstacle does not need to be expressed in the proposition, but if it is expressed, it may be related to knowledge, information, reasonable thinking or simply the participant’s belief. 5. Positive (non-neutral) attitude in connection with epistemic modality a) Han må ha reist til Paris. ‘He must have left for Paris.’ = It is not OK for me to believe anything else than that he has left for Paris. b) Han vil ha reist til Paris. ‘He will have left for Paris.’ = It is not OK for me to believe anything else than that he has left for Paris.

The utterance in (5a, b) represents simple attitude, which is to say that the speaker is not accepting any other attitude besides that the proposition Han er reist til Paris ‘He has left for Paris’ is correct. The difference between (5a) and (5b) is not related to the attitude as such. The utterance in (5a) may be characterized as a deduction, while (5b) may be characterized as a prediction. In other words, by choosing the modal verb må the speaker signals that she has some kind of evidence to base her conclusion on. By contrast, the modal verb vil is chosen when the speaker predicts something without necessarily having any evidence (or when she refers to something that is known commonly). Thus, the choice of the modal verb provides more information than barely the speaker’s attitude sensu stricto. As in the domain of non-epistemic modality, the speaker may also possess a complex attitude, as is demonstrated by the utterances in (6). 6. Complex attitude in connection with epistemic modality De bør ha reist til Paris. ‘They ought to have left for Paris.’

Complex attitude means that the speaker considers the propositions in (6) to be correct, but at the same time she signals that different points of view may be accepted as well. In practice such utterances as

130  Ugnius Mikučionis in (6) may be interpreted as representing a lower degree of confidence (certainty) compared to the utterance in (5a, b). The use of modal verbs such as kan, bør, vil, må does not in itself contain information about the reasons to believe that the proposition is true (or the nature of obstacles to reject it), other than what has been said about the difference between deduction (expressed by må) and prediction (expressed by vil). The situation is somewhat different when the modal verb skal is used, as discussed below in the subsection on evidentiality. Dynamic modality and evidentiality – modal or just modality-related domains? In this section, I will briefly discuss the status of dynamic modality and evidentiality in relation to my model of the semantics of modality. Dynamic modality Dynamic modality is usually defined as dealing with a participant’s ability and, according to some authors, willingness to perform actions or get involved in states of affairs. Examples of these subtypes of dynamic modality would be utterances like (7) and (8). 7. Ability Han kan spille piano. ‘He can (is able to, knows how to) play the piano.’ 8. Willingness Han vil spille piano. ‘He will (wants to, is willing to) play the piano.’

Some authors extend the notion to also cover enabling conditions that are external to the participant(s). This type of modality is sometimes called circumstantial modality or circumstantial possibility. It may be exemplified by the utterance in (9).

131  Norwegian Modal Verbs and Attitudinal Modality 9. Circumstantial possibility Man kan spille piano der (det fins nemlig et piano der borte). ‘One can play piano there (there is in fact a piano available over there).’

As already mentioned, I have chosen to define modality as an attitudinal category, and by attitude I mean people’s evaluation of the trustworthiness of propositions or the desirability of states of affairs to occur. So the question is how dynamic and circumstantial modality fits into my model of modality. The status of willingness seems to cause no problems. Willingness is a kind of positive attitude, and therefore utterances expressing willingness are treated as modal. In my model, willingness is a kind of non-epistemic modality. The status of ability and circumstantial possibility is less clear. If ability is taken to mean nothing else but a person’s mental or physical powers, it falls outside the range of what can be called modal (= attitudinal) meanings. However, the Norwegian verb KUNNE often is used to signal that it is OK if a state of affairs occurs or if a proposition turns out to be true without specifying why it is OK. Or, to put it in other words, the verb KUNNE is often used to express that there is no obstacle for a state of affairs to occur or for a proposition to be true. Since there is no obstacle, the speaker does not need to say anything about the nature of the obstacle(s) which might potentially prevent the state of affairs from occurring or rule out the chances that the proposition could be true. The utterance in (7) Han kan spille piano ‘He can play the piano’ may be interpreted as meaning ‘He agrees (is not unwilling, has nothing against) playing the piano’, as well as ‘He is able, knows how to play the piano’ – and ‘He may be playing the piano’. It is impossible to tell which of the interpretations was intended by the speaker unless additional information is provided. This is not to say that there is no difference between the three interpretations, or that it is impossible to tell them apart in principle. The speaker may make clear which of the interpretations she intends by lexical means, but she also may fail to provide any additional information and in so doing leave it for the hearer to choose which one of the interpretations he prefers. The fact that a person knows how to play a piano does neither require him to be agreeable nor prevent him from being agreeable to play a piano. There

132  Ugnius Mikučionis is no automatism in the relationship between ability and agreeability (positive attitude). So, it is clear that the two interpretations are distinct in principle. The one of them (ability) is not related to attitude sensu stricto, while the other one (agreeability) clearly is. Ability and agreeability may, but do not need to, coincide. Therefore it seems reasonable to claim that the ability-reading and the agreeability-reading of the verb KUNNE belong to different squares on a figure representing the semantics of the Norwegian modal verbs rather than to the same square. But those squares must be adjacent to each other, since the speaker can fail to indicate the boundary between them. The same goes, mutatis mutandis, for the relationship between epistemic and dynamic modality. The relationship between circumstantial modality and non-epistemic modality (attitude) is of the same kind. The utterance in (9) above, Man kan spille piano der (det fins nemlig et piano der borte) ‘One can play piano there (there is in fact a piano available over there)’, will unambiguously be interpreted by the hearer as an example of circumstantial modality only if the remark about availability of a piano is included. This interpretation involves no attitude. Without additional information the utterance Man kan spille piano der ‘One can play piano there’ may also be interpreted as expressing someone’s permission, i. e. attitude. Availability of a piano and someone’s permission may, but do not need to, coincide. The speaker may say explicitly whether she speaks about the availability of necessary resources or about authorities’ permission. But she may choose just to say that there is no obstacle for playing a piano, leaving it for the hearer to decide which one or both of the interpretations he chooses. Thus, it may be argued that the difference between neutral attitude on the one hand, and absence of (physical, material or any type of) obstacles on the other hand, is linguistically irrelevant, at least in the standard bokmål variety of the Norwegian language. By linguistically irrelevant I mean that the speaker does not necessarily need to express her choice vis-à-vis the mentioned difference. Returning to the question about the position of dynamic (and circumstantial) modality in my model, it seems reasonable to claim that dynamic (and circumstantial) modality is connected to epistemic and non-epistemic modality via underspecification. The attitudinal and non-attitudinal meanings of the verb KUNNE are clearly distinct, and the speaker may express

133  Norwegian Modal Verbs and Attitudinal Modality explicitly which of the meanings she intends. But the speaker may choose to fail to draw the boundary between attitudinal and non-attitudinal meanings, leaving it to the hearer to choose between interpretations. Evidentiality The modal verb skal is not only used to express the participant’s attitude. It is also frequently used in utterances where the speaker refers to someone else’s words. That usage represents one of the so-called evidential meanings. Thus, by uttering Han skal ha reist til Paris ‘He “shall” have left for Paris’, the speaker indicates that it is someone else that claims that the proposition Han har reist til Paris ‘He has left for Paris’ is true. The degree of the speaker’s own commitment to the proposition is not unambiguously shown by the modal verb in this case. Of course, it may also be shown by some other means of expression. The speaker who refers to someone else’s words may also want to express to what degree she herself is committed to the trustworthiness of the proposition, but in such a case she must choose some other means of expression. It seems that it would be “too much work” for an auxiliary verb to indicate both the source of information and the degree of trustworthiness. In a sentence which contains both the information that the speaker is reporting someone else’s words and information about the degree of trustworthiness, one has to employ two means of expression. 10. Reported proposition and degree of trustworthiness a) Indirect knowledge is indicated by a lexical expression (Hun sier), and attitude towards the trustworthiness of the proposition is indicated by the choice of the modal verb. Hun sier at han kan / bør / vil / må ha reist til Paris. ‘She says that he may / ought to / will / must have left for Paris.’ b) Indirect knowledge is indicated by means of the evidential verb skal, while attitude towards the trustworthiness of the proposition is indicated by a lexical expression. Han skal ha reist til Paris, men det tror jeg ikke noe på. ‘He is said to have left for Paris, but I don’t believe this is correct.’ Han skal ha reist til Paris, og det kan godt stemme. ‘He is said to have left for Paris, and this may well be the case.’

134  Ugnius Mikučionis Evidential skal is thus not related – not directly, at least, – to the speaker’s own attitude towards the trustworthiness of the proposition. However, the preterite form skulle may be used to indicate a lower degree of the speaker’s commitment to the trustworthiness of the proposition than the present tense form skal. Consider the two sentences in (11). 11. Evidentiality and degree of trustworthiness a) Han skal ha reist til Paris. ‘He is said to have left for Paris’ (and I say nothing about the level of trustworthiness) b) Han skulle ha reist til Paris. ‘He is said to have left for Paris’ (and I see this information as less trustworthy)

The  difference between skal and skulle can most probably be accounted for from a diachronic point of view, treating the form skulle as subjunctive of the verb SKULLE. In Modern Norwegian, however, there seem to be no grammaticalised means to express both the fact that information is reported and the degree of the speaker’s commitment to the trustworthiness of this information. In the traditional literature on modality, reported information is treated as a type of evidentiality, but there is no consensus as to whether evidentiality is to be included into the domain of epistemic modality or if it should be considered as a separate, though adjacent, domain. Since modality is defined as an attitudinal category, that is, a cate­ gory dealing with the people’s attitudes towards propositions or states of affairs, the question about the source of information is irrelevant, so to speak, for the decision whether a category is modal or not. A crucial question, however, is whether the category in question describes some participant’s attitude towards the validity of a proposition, or not. In such a perspective, evidentiality should only be treated as a modal category to the extent that it involves an evaluation of the OK-ness of a proposition. These considerations point in the direction of evidentiality being a non-modal domain, in principle. This interpretation is also supported

135  Norwegian Modal Verbs and Attitudinal Modality by the fact that a feature “reported” may, but does not need to, be combined with information about the obstacles to accept a proposition as true / false – it follows from there that evidentiality and modality are two different categories. Conclusion To sum up the proposal above, we can represent the different types of modality as in figure 3. Non-epistemic modality Positive attitude

Neutral attitude

Epistemic modality

Complex attitude

not OK unless a state not OK unless a of affairs occurs (but proposition is true there is room for (but there is room for alternative attitudes) alternative attitudes)

Simple attitude

not OK unless a state of affairs occurs

not OK unless a proposition is true

OK if a state of affairs OK if a proposition occurs (but also OK is true (but also OK if it does not occur) if it is false)

Figure 3. Overview of the types of modality.

A rough overview of the uses of modern Norwegian modal verbs in terms of the proposed model is given in figure 4 (see page 136). The real picture is further complicated by pragmatic considerations, such as the use of kan / kunne in imperative utterances with a certain amount of politeness or, on the contrary, irony and impatience. Such utterances deserve a more detailed discussion (cf. Mikučionis 2009). Another important aspect of the use of the Norwegian modal verbs is preterite (past tense) forms, used non-temporally. They have not been covered in the current article.

136  Ugnius Mikučionis

Positive attitude

Complex attitude

Simple attitude

Neutral attitude

Non-epistemic modality

Epistemic modality

bør, burde, skulle, ville

bør, burde, skulle

må (unspecified må, vil source of attitude), skal (personal or institutional source of attitude), får (approx. “have no choice”), vil (willingness) kan, kunne (unspecified source of attitude), får (“is allowed”), må (in connection with gjerne / bare)

kan, kunne

kan (dynamic / circumstantial meanings) Figure 4. Uses of modern Norwegian modal verbs.

Bibliography Andersson, Peter 2007. Modalitet och förändring. En studie av må och kunna i fornsvenska. Göteborg: Göteborgs universitet. Bybee, Joan, Revere Perkins & William Pagliuca 1994. The evolution of grammar. Tense, aspect and modality in the  languages of the  world. Chicago / London: The University of Chicago Press. Mikučionis, Ugnius 2009. Modality and the Norwegian verb. Unpublished manuscript, Vilnius University. Palmer, Frank 2001. Mood and modality. Second edition. Cambridge: Cambridge University Press.

137  Norwegian Modal Verbs and Attitudinal Modality van der Auwera, Johan & Vladimir A. Plungian 1998. Modality’s semantic map. Linguistic Typology 2, pp. 79–124. van der Auwera, Johan, Andreas Ammann & Saskia Kindt 2005. Modal polyfunctionality and Standard Average European, in: Klinge, Alex & Henrik H. Müller (eds.): Modality: Studies in Form and Function. Oakville, pp. 247–268.

The View of Blood Vengeance in Medieval Norwegian Sources1 Else Mundal University of Bergen

The picture we have of Old Norse society as a feuding society where blood vengeance flourished is essentially based on the Icelandic sagas. The sagas of Icelanders (Íslendingaso˛gur) describe a feuding society in the period before Christianization. Blood vengeance is also a theme in the contemporary sagas which describe events in the twelfth and thirteenth century – even though the violence described in this literature is normally more closely connected to the political struggle for power than to ordinary family feuds. The Icelandic sources leave no doubt, however, that blood vengeance continued to be practiced in Iceland during the Free State period and later, but the sagas of Icelanders, which idealize the pre-Christian period and the hero who defends his honour, most likely give an exaggerated picture of how common blood vengeance was. The Norwegian sources are less rich. In some cases the blood vengeance described in the sagas of Icelanders takes place in Norway, and the kings’ sagas normally describe the Norwegian condition whether the author is an Icelander, an Icelander living in Norway or a Norwegian. In the cases where the author is an Icelander, it would be reasonable to assume that he was influenced by Icelandic conditions and described blood vengeance in accordance with the Icelandic practice. It is, however, very difficult to point out differences between the two countries in the descriptions of blood vengeance in literary sources. This may be because blood vengeance was still practiced similarly in the two countries at the time the saga literature was written. At least some authors may also have been aware that the conditions under which blood vengeance 1 This article is based on a paper read at Leeds International Medieval Congress, 12–15 July, 2004. Approaching the Viking Age. Proceedings of the International Conference on Old Norse Literature, Mythology, Culture, Social Life and Language. Edited by Ērika Sausverde and Ieva Steponavičiūtė. (Scandinavistica Vilnensis 2). Vilnius University Press, 2009. https://doi.org/10.15388/ScandinavisticaVilnensis.2009.2.8 Copyright © 2019 Authors. Published by Vilnius University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

140  Else Mundal was practiced were even more similar in the time described in most of the sagas than in the period in which they were written. Blood vengeance was closely connected to honour – to the necessity of re-establishing the honour of a man or a family if it had been damaged. Blood vengeance was also closely connected to a certain kind of society, one where the central power or authority of the state was weak and the families had to rely on themselves to protect their lives and property. Christianization, which took place in Norway and Iceland at about the same time, would, over time, change ideologies with roots in the heathen culture. The opinion that all injustice should be avenged to protect one’s honour, and if necessary by killing, would come in conflict with Christian ideas such as forgiveness, humility and the commandment “Thou shalt not kill”. We must assume that from the very beginning the Church worked against blood vengeance, but with what vigour we do not know. While both Norway and Iceland were Christianized around the year 1000, and the Church was established in both countries with bishoprics in the eleventh century, only Norway was a kingdom at the time of Christianization. Iceland continued to be a Free State without an executive power until the union with Norway in 1262/64. The fact that blood vengeance continued to flourish in Iceland – though probably more in literature than in reality – has often been seen in connection with the absence of an executive power. Whether there was less blood vengeance in Norway than in Iceland in the first centuries of the Christian period cannot in fact be determined. However, as time went on the kings of Norway probably began to see the institution of blood vengeance as inconsistent with the king’s role as lawmaker and protector of law and order in his country. Therefore, if the joint forces of the Church and the king in Norway were both working against blood vengeance, we would expect a slightly different development in the two societies regarding the use and view of blood vengeance. It is therefore of interest to look for the impact of the king’s work against blood vengeance in Norwegian sources. The laws are the most valuable sources for this as it is here that we can expect to find reflections of the king’s policy as a lawmaker. People’s view of blood vengeance may also be reflected in the laws, although perhaps more indirectly. Attitudes towards blood vengeance find expression

141  The View of Blood Vengeance in Medieval Norwegian Sources in literature of different kinds. The author of Konungs skuggsjá (The King’s Mirror), from the middle of the thirteenth century, comments on the phenomenon, and blood vengeance is a theme in many Norwegian ballads which probably reflect attitudes in the late Middle Ages and later. The Old Norse language does not have a word for ‘blood vengeance’. There is only a word for ‘vengeance’, hefnd (f.), and the corresponding verb is hefna. These words are used both when a person beats someone who has beaten him first, when a person calls someone names who has abused him first, and when a person kills someone to avenge a killing of a kinsmen or an earlier serious insult. Hefnd is, however, also used as a word for ‘punishment’. Thus, in some cases it may be difficult to decide whether the laws describe the old institution of blood vengeance which was carried out in the family’s own interest to protect the family honour, or whether hefnd was sometimes in reality a death penalty which people were instructed or encouraged to execute on the king’s behalf. I will return to this question later. In the oldest layers of the Norwegian laws there are rather many cases in which a man can legally avenge himself by killing. If we take our point of departure in The Older Gulathingslaw,2 the law of the district of Western Norway which probably reflects the oldest stage found in the preserved laws, it is stated that a man is allowed to kill someone immediately, without first bringing the case to court, in the following circumstances: •  if he finds a man in bed with his wife, sister, daughter, mother, stepmother, his brother’s wife or his son’s wife (ch. 160); •  if he catches a thief stealing food from his storehouse or an animal from his cattlehouse or fold (ch. 160); •  if he, or the people he is with, are attacked, he can kill in defense, and if the people he is with are killed, he can go after the killer and take revenge by killing him (see, for example, chs. 152, 167, 171, 189, 195).3



2 In Norges Gamle Love 1. Parallels to The Older Gulathingslaw in other Norwegian laws can be found under the entry words hefna and hefnd in Norges Gamle Love 5. Parallels in the laws of the Icelandic Free State can be found under the entry word vígt in Grágás 1883. 3 Jørn Øyrehagen Sunde has discussed blood vengeance in the cases mentioned above (Sunde 2005: 56–63). He points out that blood vengeance as described here has many parallels in laws outside Scandinavia. See also Sunde 2006.

142  Else Mundal These may be regarded as examples of the right to defend one’s own life and property, the lives of people a man was expected to defend, and the lives and honour of the kinswomen. These cases are, in principle, different from blood vengeance as practiced in a feud. The examples above describe immediate and spontaneous reactions to a killing. In cases where men had to run after the killer, it is obvious that revenge did not take place at the location of the crime, and occurred only after some time. The Norwegian laws do not say how much time the avenger had to carry out the act of revenge, but if we look at the parallels in Icelandic laws, in most cases revenge had to be carried out within the same day. Only in the most serious cases – killing and rape – could the offended part in some special cases wait longer, but he had to take revenge before the next Thing. In these cases the avenger had time to prepare himself and plan the revenge. This is typical of blood vengeance as practiced in a feud. However, as the feud is described in saga literature, there was no time limit for revenge. Blood vengeance could take place many years after the killing or insult that caused the act of revenge. There is another even more fundamental difference between the cases of legal blood vengeance mentioned above and those in a feud. In the cases above, blood vengeance was directed towards the criminal himself. In a feud, revenge would not necessarily be directed against the offender, but could be directed against anyone of his kinsmen, preferably the best men of the family. There was also a case, other than those mentioned above, where a man, according to the older layers of the laws, could legally defend himself with blood vengeance. The Older Gulathingslaw (ch. 196) gave a man the right to avenge himself with blood vengeance if he was accused of having given birth to children, of having been the sexual partner of another man (in which he played the sexual role of the female), or had been compared to a female animal of any kind or to a whore (ch. 196).4 Insulting words were normally not reason enough for blood vengeance: skal orð orðs hefna, ‘[insulting] words shall be avenged

4 Parallels to The Old Gulathingslaw in other Norwegian laws can be found under the entry word fullréttisorð in Norges Gamle Love 5. Parallels in the laws of the Icelandic Free State can be found under the entry word vígt in Grágás 1883.

143  The View of Blood Vengeance in Medieval Norwegian Sources by  [­insulting] words’, as it is written in The  Older Frostathingslaw5 (X, ch. 35). However, serious insults did result in a fine, and only especially libellous words gave a man the right to avenge himself with blood vengeance. Nor do the Norwegian laws in this case say how soon the offended man must react to the insulting words in order to ensure that the killing would be legal, but according to Icelandic laws he had to react before the next Thing. As in the other cases where a man could legally defend life and honour by killing in revenge, blood vengeance had to be directed against the man who was guilty of the crime, not against members of his family. Blood vengeance, as it was practiced in a feud, is in fact not directly addressed in Norwegian laws – nor in Icelandic laws either for that matter – before the middle of the thirteenth century. Of course any killing, even those looked upon as blood vengeance, but not legal vengeance of the type mentioned above, could be handled according to the law: the case was brought to court by the offended party, the killer and his family were sentenced to pay fines, and the killer, and perhaps some of his helpers, could be outlawed. Then the case would be settled, but only for awhile; the next revenge would start the same procedure over again, only with the two families in opposite positions. As time went on, illegal revenge could be difficult to handle according to the law since some – or even many – of the men involved in the feud would be outlawed, which meant that they could not bring a case to court and they would not be entitled to compensation if attacked. Judging from the sagas of Icelanders, which best describe how feuds developed, we get the impression that sometimes blood vengeance was handled according to the law but that sometimes the case was not brought to court. To bring a case to court was not an option if the person who had been killed in revenge was already outlawed. The offended family had nothing to gain by following legal procedure, and if they wanted to continue the feud they had to do so outside the law. The situation must have been very much the same in Norway as in Iceland, but as time went on and the power of the king grew stronger, it may have been more difficult to get away with illegal revenge in Norway.

5 In Norges Gamle Love 1.

144  Else Mundal An attempt to limit blood vengeance can be observed in the Nor­ wegian laws from the middle of the thirteenth century onwards. In the younger layers of The Older Frostatingslaw, probably from the last years of King Hákon Hákonarson’s reign, the king appears as a legislator. The new preface to the law takes the form of a letter from the king to all Norwegians, and here, as in ch. 8 of the first section, we see for the first time a prohibition against taking revenge on someone other than the killer. The king says that killing the best men from a family in revenge instead of the killer himself has been a bad custom for a long time in this country, and it is now prohibited. The same prohibition in more or less the same wording is also found in the so-called Hákonarbók,6 ch. 20. This law was meant for Iceland, but some chapters mention that the provision in question had already been made law in Norway. The law is therefore primarily a source for new ideas concerning law and justice in circles around the Norwegian court. One innovation in this law is that only the man who committed the crime had to pay fines, not his kinsmen (ch. 42). This legal principle is also repeated in the new law for the whole country, The Landslaw7 (Landslo˛g), given by King Magnús lagabœtir, ‘the lawmaker’, in the 1270s.8 While this innovation was a relief for the kinsmen it led to financial problems for the criminal. Finally, the right of kinsmen to seek revenge by killing was considerably restricted in this period. In the laws from King Hákon Hákonarson’s time, kinsmen still had the right to take revenge if a killer or rapist refused to pay his fines (The Older Frostathingslaw I, chs. 5 and 6), and Hákonarbók, ch. 20, states that if a man kills without reason he is utl gr oc ugildr … b ði konungi oc fr ndom, ‘outlawed and has no legal rights … as regards both the king and the family [of the dead]’, which means that not only the king but also the kinsmen of the dead man had the right to kill him. To kidnap or run away with a man’s wife was another crime for which kinsmen retained the right to kill; according to the law it was more serious to steal a man’s wife than a man’s cow. Such men are obota menn b ði fire konungi oc karle dr pir oc deyddir hvar sem þ ir

6 In Norges Gamle Love 1. 7 In Norges Gamle Love 2. 8 The provisions are found in section X at the end of the law.

145  The View of Blood Vengeance in Medieval Norwegian Sources verða stadder, ‘men for whom no compensation shall be paid as regards both the king and [other] men, they can be killed and put to death wherever they are’ (Hákonarbók, ch. 20). The phrase b ði fire konungi oc karle indicates that not only the king and his men were allowed to kill the criminal in such cases, but other men as well. The phrase that a man is dr pr b ði fire konungi oc karle, ‘that a man can be killed both by the king and other men’, is also found in The Landslaw (IV chs. 3 and 4). Here it is stated that a man has the right to kill to defend his property and his kinswomen, but in this law punishment by the king has, in principle, replaced blood vengeance (IV, ch. 16). It is obvious that King Hákon Hákonarson and King Magnús lagabœtir played a very active role in reducing the practice of blood vengeance in Norway. The question is whether the earlier kings did not try or simply did not succeed in bringing blood vengeance under their royal control. At first glance it may seem as if the people of Norway were even more eager for revenge than Icelanders were. The Older Gulathingslaw (ch. 186) states that a man cannot receive compensation more than three times if he does not avenge himself in between. If the king wanted to reduce the use of revenge and the number of killings – which was the king’s explicit policy in the thirteenth century – one would think that revenge as a prerequisite for receiving compensation would have given the freeholders’ ideas. On the other hand, the free farmers were normally more interested in the right to avenge themselves than in a duty to avenge themselves; and seen from the king’s point of view, vengeance could have a deterrent effect on negative elements in society. If we compare the laws from the Free State of Iceland and the kingdom of Norway in the period before the middle of the thirteenth century, which give men the right to protect themselves and their interests by killing in more or less the same cases, there is one interesting difference. In the Norwegian laws men are not only granted the right to take revenge, they are in some cases strongly encouraged or even instructed by the law to pursue a man and kill him. If a man is killed on a ship, “then it is good if he is avenged or thrown over board”, The Older Gulathingslaw (ch. 171) states. The same law also states that if a man kills someone in a group and runs away to the forest, the rest of the group has to run after him; and it is implied that they should kill him. The Older

146  Else Mundal Frostathingslaw IV, ch. 9, commands that if a man is killed (ho˛ggvinn) at the Thing, all men should run after him to the forest. The next chapter states that if a man is wounded at the Thing, all free men ought to run after the criminal. Chapter 13 in the same section of the law stipulates that if a man is wounded, wherever he is, all free men should run after the evil-doer. In all these cases it is implied that everyone should run after him to kill him. Another interesting request for the family of a man who has been killed to pay back with blood vengeance is found in The Older Fros­ tathingslaw IV, ch. 33, and in Hákonarbók, ch. 22. According to these laws the kinsmen of the dead are allowed to kill a woman who has killed her own husband or caused his death. In Old Norse culture, women were normally not the target of blood vengeance. To kill women and children would have been shameful. If a woman was guilty of killing, revenge would most likely have been directed against one or more male members of her family. The provision could be seen as a request to kill the evil-doer instead of an innocent man from the woman’s family. There is, however, one law in particular which indicates that earlier kings at least tried and had limited success in taking over the families’ old right to settle their own affairs. In all Norwegian laws a serious crime against a person was not only a crime against the person in question and his or her family, but was also a crime against the king, and the king was entitled to compensation. In The Older Bjarkeyréttr,9 the law for the towns, kinsmen and their right to receive compensation when a relative has been killed are mentioned only a few times. According to this law, income from fines was normally divided between the king and the men of the town who had more or less replaced the family. In this law we also find a very interesting example of how execution on the king’s behalf could develop out of blood vengeance. In The Older Bjarkeyréttr (II, ch. 13) the kinsmen of a man who has been killed are instructed to kill the murderer, but if the kinsmen were not present, one of the king’s civil servants had to do the job. The cases in The Older Gulathingslaw and The Older Frostathingslaw in which the law instructs or encourages men to kill an evil-doer, could also be seen as examples

9 In Norges Gamle Love 1.

147  The View of Blood Vengeance in Medieval Norwegian Sources of how kings used the old custom of blood vengeance to promote their own interests.10 When a man killed a criminal who had offended him or members of his family in revenge he was acting in his own interest, but in many cases it was also in the interest of the society to eliminate criminals. The step from blood vengeance to punishment was a short one in these cases. At least in some cases the laws of the towns can be seen as supplementary to the laws of the surrounding districts. The differences between the two laws may therefore be smaller than a comparison of the texts seems to indicate. Nevertheless, these differences seem to imply that the king had greater power in the towns than in the rural districts. The lack of control in the districts is in fact clearly reflected in the younger part of The Older Frostatingslaw, given by King Hákon Hákonarson. In section I, ch. 12, of this law the king complains about outlawed men living in the countryside, even protected by the king’s own civil servants. The towns, which according to later written sources were established by the kings and were their main residences, may also have been bridgeheads for the kings’ power and new ideas promoted by the king and the Church. The absence of kinsmen and extended families combined with the presence of the king in the new towns made this possible. If the differences between Norwegian and Icelandic laws mentioned above developed as a result of the kings’ interference, this could point to the kings’ struggle for control as the cause, rather than new ideas. This is what we would expect in the first centuries after Christianization. The instructions or encouragements to kill the evil-doer which are found in the older layers of the laws are, however, in accordance with ideas expressed in the new laws from the middle of the thirteenth century. 10 There are, however, a few provisions in The Older Frostathingslaw (IV, chs. 50, 51, 52) which cannot have resulted from the king’s policy. These provisions state that if the king, the earl or the lendr maðr kills a man without reason, the farmers should kill him. It is very difficult to say when these provisions found their way into the written law and what the background for them might have been. One possible explanation could be that the king had to accept this so that the farmers would accept the duty of killing criminals in cases where they had no personal interest. For a discussion of these provisions, see Bagge 2005.

148  Else Mundal In the middle of the thirteenth century a literary work appeared which makes an interesting source for a study on the views of blood vengeance. The anonymous work Konungs skuggsjá (The Kings Mirror) was written in circles closely connected to the  Norwegian king. The author regards vengeance as a duty a man has to undertake when necessary, but recommends moderation in the execution of revenge (Holm-Olsen 1983: 66). He does not question the morality of vengeance; what he finds objectionable is that the fines and the vengeance are the same whether the man who has been killed was a good and wise man or an evil and stupid man (Holm-Olsen 1983: 54).11 The ban on taking revenge on anyone other than the killer himself and the command that the criminal should pay his fines alone are innovations that could be seen, at least partly, as consequences of the Christian religion. According to Christian beliefs sin was an individual problem. When the Norwegian king argued that it was a bad custom to kill a man who had done nothing wrong, many people probably agreed with him in principle. Whether it really had been more common in Norway than elsewhere to take revenge on the best man in a family, as King Hákon Hákonarson claims, cannot be determined from the sources, and whether the new provisions actually made a difference is also very hard to say. According to King Hákon Hákonarson’s new preface to The Older Frostathingslaw, the main reason for restricting the use of blood vengeance was to reduce the number of killings. Whether the number of killings was actually reduced or not we cannot say, but we do know of approximately three hundred cases of manslaughter or murder in Norway in the two last centuries before the Reformation, and this is probably only the tip of the iceberg. Criminals are to be found among all classes of society, the men of the Church make no exception.12 It is often difficult to see from the sources whether a killing was regarded as revenge or not. However, there are some typical examples of blood vengeance, and a few of them 11 On the view of punishment and revenge in Konungs skuggsjá see Bagge 1987, especially chapter II, “The King as Judge”; Bagge forthcoming, especially the chapter “Justice, Law and Power”. 12 See Jørgensen & Saletnich 2004.

149  The View of Blood Vengeance in Medieval Norwegian Sources clearly illustrate how the Christian culture and the old demand of blood vengeance lived side by side. A diploma from Skien (Telemark) from 133813 tells how a man had been fatally wounded by an enemy. The brothers of the dying man ran to find a priest, but on their way to the priest they met the man who had wounded their brother and they stopped to kill him in revenge. Saving their brother’s soul and protecting their honour were seen as equally important. An interesting regulation which also concerns blood vengeance was made by the bishop of Oslo in the year 1395.14 The regulation is in the form of a diploma to the people from the district of Telemark, and the bishop accuses the people of this district of committing more killings than any other district in Norway. In relation to blood vengeance, it is interesting that the bishop accuses the people of this district of taking revenge after having accepted compensation, and he threatens them with “heluitis pinu medh di flinum si luum”, ‘torment in Hell together with the Devil himself ’.15 This indicates that blood vengeance continued to flourish, at least in this particular district, and that the pressure from the authorities to settle a case after having received compensation was greater than the farmers’ willingness to abstain from vengeance, which some people might still have considered the most honourable resolution. Killing as a reaction to insulting words is typical in many of the murder cases described in Norwegian diplomas.16 In some cases the insults might share some traits with the type of insults found in the oldest layers of the laws which gave a man the right to avenge himself by killing. But if the killings in reaction to insults described in Norwegian diplomas can be regarded as examples of blood vengeance, in most cases they are examples of extreme proportions. Even so, in many cases minor insults seem to have led to killings. Blood vengeance is also the theme of many ballads. The attitude towards blood vengeance in this literature, which probably reflects 13 Diplomatarium Norwegicum 1: 196f. 14 The regulation is preserved in a younger copy and is printed in Taranger (ed.) 1912: 328–334. 15 Taranger (ed.) 1912: 329. 16 See Solberg 2003a, especially chapters 7 and 8.

150  Else Mundal the ideology of the late Middle Ages both in Norway and in the rest of Scandinavia, is interesting in that it demonstrates that there was not only one view of blood vengeance, but many. The tragedy of blood vengeance is the theme of a ballad like Hemnarsverdet17 (The Sword of Vengeance). In this ballad the sword takes control and kills everyone, reflecting an attitude towards blood vengeance that is highly negative. A similar plot is found in the ballad Mindre Alf.18 Here the avenger manages to stop the sword by mentioning the name of God. In other ballads, however, blood vengeance is more or less idealized. In the ballad Ivar Elisson,19 the mother who is afraid that her son will be killed when taking revenge says that it is better to live with shame than to lose one’s life. Nevertheless she is very pleased when vengeance is carried out. Here the attitude towards blood vengeance is ambiguous. The most common plot in these ballads is of a son who takes revenge by killing the man who murdered his father, as, for instance, in the ballad Tiarmann i Stokkholmen.20 In such cases blood vengeance is seen as a necessary and honourable thing to do. In some ballads it is the daughters who carry out revenge, as in the ballad Sigrid and Astrid. 21 Here blood vengeance is highly idealized. The fact that daughters take revenge on the man who has killed their father emphasizes the necessity of vengeance. However, the tragedy of blood vengeance is often focused upon, even in ballads that describe revenge as a necessity. In some ballads a woman is in the unfortunate position of having lover or husband who has been killed by her father or brother and she turns against her own family and kills a close relative in revenge, as in the ballads Herr Hjelmen22 and Far og dotter23 (Father and Daughter). Finally, we have ballads which look at blood vengeance from a humorous point of view. The ballad Kjerringa vil skrifte24 (The Old

17 Printed in Olav Bø & Svale Solheim (eds.) 1958. 18 Printed in M. B. Landstad (ed.) [1853] 1968. 19 Printed in Olav Bø & Svale Solheim (eds.) 1958. 20 Printed in Olav Bø & Svale Solheim (eds.) 1959. 21 Printed in Olav Bø & Svale Solheim (eds.) 1959. 22 Printed in Olav Bø & Svale Solheim (eds.) 1959. 23 Printed in Olav Bø & Svale Solheim (eds.) 1959. 24 Printed in Solberg (ed.) 2003b.

151  The View of Blood Vengeance in Medieval Norwegian Sources Woman Wants to Confess) tells about an old woman who takes revenge by killing a nobleman who has stolen her porridge. She goes first to the bishop to confess her sin, but he sends her to Rome to confess to the pope who seems to approve of her deeds. One of the themes of this ballad is vengeance that is extreme in its proportion, and this ballad can be read as a humorous comment on killings as reactions to minor insults.25 The essentially different attitudes towards blood vengeance in sources from the late Middle Ages indicate that the views among common people were changing in this period. The tragedy of blood vengeance, especially for women, may also be a theme in Eddic poetry and saga literature. However, the ballads express this tragedy more clearly; honour can be bought at a too high price. It is also worth noting that the kind of blood vengeance which King Hákon Hákonarson looked upon as absolutely reprehensible – that which had someone other than the killer as a target – is no longer a theme in the ballads, and probably was without defenders in the late Middle Ages. 25 Ballads of this type are normally late, but this ballad must be medieval and was written down in Denmark as early as in the 16th century.

Bibliography Bagge, Sverre. 1987. The Political Thought of The King’s Mirror. Odense: Odense University Press. ———. 2005. Kirken, bøndene og motstandsretten i Norge i middelalderen. Historisk tidsskrift 84, 3, 385–410. ———. Forthcoming. From Viking Stronghold to Christian Kingdom. State Formation in Nor­way c. 900…1350. København: Museum Tusculanum Press. Bø, Olav & Svale Solheim (eds.). 1958. Folkeviser I (Norsk folkedikting VI). Oslo. Bø, Olav & Svale Solheim (eds.). 1959. Folkeviser II (Norsk folkedikting VII). Oslo. Diplomatarium Norwegicum 1. 1847. Chr. C. A. Lange & Carl R. Unger (eds.). Christiania. Grágás 1883 = Grágás. Stykker, som findes i det Arnamagn anske Haandskrift Nr. 351 fol. Skálholtsbók og en R kke andre Haandskrifter. Vilhjálmur Finsen (ed.). København 1883, reprint Odense 1974. Holm-Olsen, Ludvig (ed.). 1983. Konungs skuggsjá (Norrøne tekster nr. 1). Oslo.

152  Else Mundal Jørgensen, Torstein & Saletnich, Gastone. 2004. Synder og pavemakt. Botsbrev fra Den Norske Kirkeprovins og Suderøyene til Pavestolen 1438–1531. Stavanger. Landstad, M. B. (ed.). [1853] 1968. Norske Folkeviser. Oslo. Norges Gamle Love 1. 1846. R. Keyser & P. A. Munch (eds.). Christiania. Norges Gamle Love 2. 1848. R. Keyser & P. A. Munch (eds.). Christiania. Norges Gamle Love 5. 1895. Gustav Storm & Ebbe Hertzberg (eds.). Christiania. Solberg, Olav. 2003a. Forteljingar om drap – kriminalhistorier frå seinmellomalderen. Bergen. Solberg, Olav (ed.). 2003b. Norske folkeviser. [Oslo]. Sunde, Jørn Øyrehagen. 2005. Speculum legale – rettsspegelen. Bergen. Sunde, Jørn Øyrehagen. 2006. Internasjonaliseringa av retten i mellomalderen som forskingsutfordring. In: Jørn Øyrehagen Sunde (ed.) Rettstekstar i mellomalderen – Idé og praksis (Rettshistoriske studier nr. 17. Institutt for offentlig retts skriftsrie, nr 6/2006). Oslo. Taranger, Absalon (ed.) 1912. Norges Gamle Love (Annen række) 1388–1604, 1. Christiania.

Far och son-motiv i nordisk hjältediktning och saga Agneta Ney Högskolan i Gävle

Vad innebar det att vara man under vikingatid och medeltid? Hur formades den manliga identiteten? I den nordiska hjältediktningen med flera andra skriftliga källor gestaltas Sigurd Fafnesbane som en idealman under vikingatid och medeltid. Mest omtalad är Sigurd i egenskap av drakdödare, och att döma av källorna var han gynnad av gudarna, älskad av kvinnorna och beundrad av männen. Det kan därför vara av intresse att studera den litterära tradition som gestaltar Sigurd, i synnerhet vilka homosociala band som förefaller ha haft stor genomslagskraft under vikingatid och medeltid. I föreliggande studie skall far och son-motiv inom Sigurdstraditionen studeras, bland annat Sigurds relationer till sin far respektive hans egen roll som far, men även far och son-relationer i tidigare generationer av völsungaätten. Föreställningar om ättens betydelse, patriarkalism, släktlojalitet respektive släktkonflikter ges stort utrymme eller till och med utgör centrala teman i hjältediktning och saga. Generellt sett hade ätten en viktig uppgift i att föra ryktet och namnet vidare, helst genom söner. Den genealogiska betydelsen av vikten av att ha en son belyses bland annat av följande strof ur Hávamál. Den anger att även om en son lät vänta på sig var denne viktig för att fadern skulle kunna få en minnessten efter sig:1 Sonur er betri þótt sé síð of alinn 1 ”Hávamál”, i Eddukv ði, ed. Gísli Sigurðsson, Reykjavík 1998 (Eddukv ði), strof 72, s. 33, (”Havamål”, i Den poetiska Eddan, översättning av Björn Collinder, 2. uppl. Stockholm 1964, s. 60). Niðr, m. kan översättas med ’släkting’ och avser närmaste manlig sådan, jfr Fritzner, Johan, Ordbog over Det gamle norske Sprog I–III, Kristiania 1886–1896, s. 820. Approaching the Viking Age. Proceedings of the International Conference on Old Norse Literature, Mythology, Culture, Social Life and Language. Edited by Ērika Sausverde and Ieva Steponavičiūtė. (Scandinavistica Vilnensis 2). Vilnius University Press, 2009. https://doi.org/10.15388/ScandinavisticaVilnensis.2009.2.9 Copyright © 2019 Authors. Published by Vilnius University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

154  Agneta Ney eftir genginn guma. Sjaldan bautarsteinar Standa brautu nær nema reisi niður að nið. En son är bättre fast sent han blev född, när mannen ej längre levde; sällan står där vid stigen en bautasten, om ej frände har rest den över frände.

Att betydelsen av en son var stor, i synnerhet inom högättade familjer, märks även i eddadikten Rígsþula. I de strofer som handlar om den högättade familjen är värt att notera att dikten endast anger att det föds söner. När släktskap, arv och egendom framhålls hänvisas enbart till en manlig sida.2 Framhållandet av betydelsen av en son är dock inte entydig i eddadiktningen. Som kontrast citeras här att en ung son inte alltid var så pålitlig:3 akri ársánum trúi engi maður né til snemma syni, veður ræður akri en vit syni, hætt er þeirra hvort. tidigt sådd åker må ingen man tro ej heller för snart sin son:



2 ”Rígsþula”, i Eddukv ði, strof 39, s. 392 (”Sången om Rig”, strof 42, s. 152), Breisch, Agneta (numera Ney), Frid och fredlöshet. Sociala band och utanförskap på Island under äldre medeltid, Acta Univ. Ups., Stud. Hist. Ups. 174, Uppsala 1994, s. 73 ff. 3 ”Hávamál”, i Eddukv ði, strof 88, s. 36 (”Havamål”, s. 62). I eddadiktningen relateras oftast till den manliga släktlinjen, med hänvisning till fadersätten. Ett sätt att göra detta är att nämna faderns namn, exempelvis: ”Sigurður eg heiti, Sigmundur hét minn faðir […]” (Sigurd heter jag, son av Sigmund […]), se ”Fáfnismál”, i Eddukv ði, strof 4, s. 232 (”Fafnesmål”, s. 209).

155  Far och son-motiv i nordisk hjältediktning och saga för åkern råder vädret, men vettet för sonen, bägge två äro farofyllda.

Völsungaättens anfader, Völsung, får tio söner och en dotter. Den äldste av sönerna är Sigurds far Sigmund. Völsung och sönerna skildras som ett kollektiv och omnämns ofta tillsammans, exempelvis ”kung Völsung och hans söner” eller ”kungen och hans söner”. Sigmund hör till de hjältar som Völsunga saga uppehåller sig vid. Han skildras som en överlevare med ofantlig styrka, mod och list. Bland annat undkommer han med livet i behåll efter att ha bundits fast i en stock i skogen tillsammans med sina nio bröder. En varghona kommer nio nätter i följd och dödar alla utom Sigmund. Inför den tionde natten får Sigmund hjälp av sin tvillingsyster Signy. Sigmund har genom systerns försorg fått honung som han smörjer in sitt ansikte med. När varghonan kommer fram till honom på natten känner hon honungsdoften och börjar slicka honom i ansiktet. Sigmund biter då varghonan i tungan och håller emot så hårt att när vilddjuret försöker slita sig loss dör det.4 Sigmund håller sig därefter gömd i skogen för att bida sin tid för att kunna hämnas på sin faders baneman, kung Siggeir, som är Signys make. Signy, som också vill hämnas sin fars död, byter en dag skepnad med en trollkvinna och söker upp Sigmund i skogen. Han attraheras av henne och sagan berättar att hon stannar hos honom i tre nätter. Därefter föds deras gemensamme son Sinfjötle. Sagan ger inte besked om huruvida Sigmundr någonsin insett att det var systern Signy som besökte honom i hans grotta. Sinfjötle omtalas som stor, stark, vacker och att han påminde mycket om völsungarna. Signy prövar Sinfjötles karaktär genom att sy ihop hans ärmlinningar med hans hud. Det hade hon gjort med sina äldre söner. Dessa hade klagat högljutt, men det gör inte Sinfjötle. När han inte ens är tio år fyllda sänder hon honom till faderns gömställe i skogen. Då är det fadern som prövar sonens karaktär

4 ”Völsunga saga”, i Fornaldar sögur Norðurlanda, fyrsta bindi, ed. Guðni Jónsson, (”Völsunga saga”). Till den svenska översättningen, Völsungasagan, övers. Inge Knutsson. Inledning Staffan Bergsten, Lund 1991, hänvisas inom parentes: (”Völsunga saga”, kap. 5, s. 116 (s. 35), kap. 5, s. 117 (s. 36), kap. 5, s. 119 (s. 38), Ney, Agneta, Drottningar och sköldmör. Gränsöverskridande kvinnor i medeltida myt och verklighet ca 400–1400, Möklinta 2007 (Ney 2007), s. 100 ff.

156  Agneta Ney genom att låta honom baka en deg. Sigmund frågar därvidlag om det inte var något speciellt med mjölet. Sinfjötle säger att det kan funnits något djur i det, men att det hade han i så fall bakat in. Då skrattar Sigmund och avslöjar att det fanns en mycket giftig orm i mjölet.5 Sigmund och Sinfjötle gestaltas i en nära relation; ett tydligt far och son-motiv. Fadern prövar sonen om han duger något till, och Sinfjötle visar sig vara förslagen, munter och orädd. De lever även som vargar i skogen. Ylande dödar de män som de möter, och en dag när de är i vargskepnad råkar Sigurd bita sin son i strupen. Han ber då enligt Völsunga saga trollen om hjälp för att lösa dem ur vargskepnaden. Omsorgen om sonen märks även då Sigmund lägger ett sårläkande blad på sin sons hals.6 Sigmund gifter sig med Borghild och får med henne två söner, Helge Hundingsbane och Hamund Sigmundsson. Helge skildras som en krigare i brynja med vass blick redan som nyfödd. Relationen mellan far och son antyds som nära. Sigmund avbryter en stridskamp för att bege sig till den nyfödde och enligt dikten ge honom en lökväxt, en symbolisk gåva.7 Borghild tål dock inte sin styvson Sinfjötle, eftersom denne dödat hennes broder i en strid. Vid ett gästabud är Sigmund ouppmärksam på att hon planerar att mörda Sinfjötle genom att lägga gift i hans öl. Sinfjötle misstänker att ölet är förgiftat och antyder detta till fadern, som viftar bort det hela genom att uppmana sin son att sila det genom skägget. När giftet verkat och Sinfjötle dör är Sigmund själv nära att dö av sorg. Sagan berättar hur han tar sin döde son i famnen och går till skogs. När han kommer fram till en fjord får han se en man i en båt. Mannen frågar om Sigmund vill att han skall forsla den döde över fjorden. Till detta säger Sigmund ja. Eftersom båten är så liten finns inte plats för dem alla tre, så Sigmund skall gå runt fjorden. Men då båten försvinner utom synhåll återvänder Sigmund hem.8 Han förskjuter då Borghild och gifter senare om sig med Hjördis.9

5 ”Völsunga saga”, kap. 7, s. 133 f. (s. 122), Ney 2007, s. 101 f. 6 ”Völsunga saga”, kap. 8, s. 124 (s. 46). 7 ”Helgakvíða Hundingsbana I”, i Eddukv ði, strof 6–8, s. 165 (”Det första kvädet om Helge Hundingsbane”, s. 168 f.). 8 ”Völsunga saga”, kap. 10, s. 133 f. (s. 57). 9 ”Völsunga saga”, kap. 10, s. 134 (s. 57), kap. 11, s. 135 (s. 58), Ney 2007, s. 102.

157  Far och son-motiv i nordisk hjältediktning och saga Den utförliga historien om Sigurd och Sinfjötle som Völsunga saga berättar, saknas i eddadiktningen. I den sägs inte heller direkt att de är far och son. Sinfjötle omnämns i Helgakviða Hundingsbana I som ”Siggeirs styvson”, samt att han var van vid vargens ylande ute i skogen. En anspelning på att denne levt i skogen som varg görs således, men utan att Sigmund omtalas.10 Andra källor som nämner Sigurd och Sinfjötle är skaldediktningen, närmare bestämt Eiríksmál. Den dikten berättar om Eiríkr blóðøx död (år 954) och att han vid ankomsten till Valhall tas emot av Sigurd och Sinfjötle som nämns tillsammans, men utan att något släktband anges.11 Släktbandet far och son uttrycks dock tydligt i Edda Snorra Sturlusonar.12 Till skillnad från de västnordiska källorna, omnämns Sigmund och Sinfjötle (Fitela) i Beowulf som morbror och systerson, en relation som för övrigt skildras mycket positiv för ett framgångrikt krigarskap.13 He related everything that he had heard men say of Sigemund, his deeds of valour, many untold things, the struggle of the son of Wæls, his wanderings far and wide, the feuds and treacheries – things that the sons of men know nothing of save Fitela (who was) with him, when he, the uncle, would tell something of such a matter to fadern his nephew, as they had always been friends in need in every struggle, and had felled with their sword large numbers of the race of monsters. Allting förtäljer han som han har hört sagas 10 ”Helgakvíða Hundingsbana I”, i Eddukv ði, strof 41, s. 172 (”Det första kvädet om Helge Hundingsbane”, strof 43, s. 172). 11 Jónas Kristjánsson, Eddas and Sagas. Iceland’s Medieval Literature, transl. Peter Foote, Reykjavík 1988, s. 95 f. 12 ”Skaldskapens språk”, i Snorres Edda, översättning av Karl G. Johansson och Mats Malm, Stockholm 1991, s. 151. 13 Beowulf and the Finnsburg fragment. A translation into modern English Prose by John R. Clark, London 1911, s. 50–51 (Beowulf), Beowulf, övers. Björn Collinder, Stockholm 1955, strof 874–885, Beowulf, ed. George Jack, Oxford 1994, vers 875, ”Sigemunde”, vers 897, ”Fitela”, not 877, s. 79.

158  Agneta Ney om Sigmunds öden, om många stordåd och sällsamma ting, välsungens vigrön och vida färder, arvfejd och nidverk, som ingen man rätt vet utom han och med honom Fjotle, ty slikt hade Sigmund för systersonen förtalt ett och vart – de voro ju städse i alla slags vigverk vapenbröder.

Sigmund får i sitt dödsögonblick veta att Hjördis väntar barn, något som är väsentligt för framhävandet av genealogin. Sigmund ges nämligen tillfälle att yttra sig om sin sons framtid (att det skulle bli en son tas för givet, min anm.). Sårad på slagfältet säger Sigmund till Hjördis att deras son kommer att bli berömd och dessutom den främste i völsungaätten: ”Þú ferr með sveinbarn ok fæð þat vel ok vandliga, ok mun sá sveinn ágætr ok fremstr af várri ætt.” ([…] du bär ett gossebarn under ditt hjärta och kommer att föda det tryggt och säkert, och den pojken kommer att bli berömd och bli den främste i vår släkt.)14 Sigurd föds således utanför släktkollektivet, vid kung Hjalpreks hov i Danmark, dit hans mor kommit efter makens död.15 14 ”Völsunga saga”, kap. 12, s. 137 f. (s. 61). 15 Enligt det äldsta germanska namnskicket var det mycket vanligt med en kontinuitet som en namngivningsprincip, det vill säga att en del av faderns namn återkom i barnets, som fadersnamnet Sig-mundr i förhållande till ­Sig-urðr. Tanken bakom det slags namngivning var att man ville befästa den genealogiska kontinuiteten och överföra värdefulla egenskaper från far till son, se Janzén, Assar, ”De fornvästnordiska personnamnen”, i Personnamn, utg. av Assar Janzén, Stockholm 1947, s. 22–186. I ”Völsunga saga” är detta slags namngivning sällsynt. Att samma förled ges till tvillingarna Síg-ny och ­Sig-mundr är dock ett annat exempel på variation av namngivning med samma förled.

159  Far och son-motiv i nordisk hjältediktning och saga Sigurds avsaknad av fadern märks i det samtal som han för med draken Fafner. Först säger Sigurd att han är både moderlös och faderlös, men sedan avslöjar han vem han är och vem som är hans far. Fafner kallar honom för ”inn fráneygi sveinn” (’flamögde sven’; återigen en anspelning på vikten för en hjälte och krigare av att ha en speciell blick) och tillstår att Sigurds far var hård och att ett sådant mod är medfött.16 Uppenbart är att Völsunga sagas mytologiska del har en annan fadersdiskurs än den senare höviska delen: den kollektiva karaktären som karaktäriserar Völsung och hans söner, det ömsinta bandet mellan Sigmund och Helge, men framför allt den nära relationen mellan Sigmund och Sinfjötle. I Völsunga sagas höviska del är far och son-relationen oviktig. Kung Gjuke håller sig i bakgrunden medan hans söner skildras som aktiva och agerar utan faderns närvaro. Sigurd som far är tämligen osynlig och kommenteras aldrig direkt i litteraturen. Han får två barn, först en dotter med sin trolovade Brynhild, och senare en son med sin maka Gudrun. När dottern, Aslaug, är född, säger Brynhild till sin fosterfar kung Heimir att flickan skall uppfostras hos honom. För Sigurd själv förefaller faderskapet till dottern vara okänt, och någon relation till sonen skildras över huvud taget inte. Det enda som publiken får veta är att sonen är tre år när Brynhild dräper honom efter att Sigurd dödats av sina svågrar. På sin dödsbädd yttrar Sigurd en föraning om sin sons öde och därmed bekräftar han även sonen som sin legitime arvinge:17 Á eg til ungan erfinytja, kann-at hann firrast úr fjándgarði, þeir sér hafa svárt og dátt, enn nær numið nýleg ráð. 16 ”Fáfnismál”, i Eddukv ði, strof 2—5, s. 231 f. (”Fafnesmål”, s. 209). 17 ”Völsunga saga”, kap. 27, s. 178 (s. 113), kap. 30, s. 188 (s. 125), kap. 31, s. 194 (s. 132). Efter Sigurðrs död föder Guðrún en dotter, Svanhíldr, Völsunga saga, kap. 31, s. 193 (s. 131), ”Sigurðarkviða in skamma”, i Eddukv ði, strof 26, s. 274 (”Det korta Sigurdskvädet”, s. 243).

160  Agneta Ney Alltför ung är arvingen min, han tar sig ej ut ur ovänners gård; till egen skam och skada ha de på sistone ändrat sitt sinnelag.

Den förändrade fadersrollen i Völsunga saga är intressant på flera sätt, inte minst för att den så uppenbart följer sagans indelning i en mytologisk och en hövisk del.18 Att Sigurds fadersroll i den höviska delen är så otydlig är trots allt logisk med tanke på de drag av riddarkultur som den delen uppbär samt att Sigurd har fullt upp med sig själv i något som kan beskrivas som en individualiseringsprocess.19 Dessutom är söners kollektiva hämnd efter en fars död inte längre ett motiv i den höviska delen, däremot bröders kollektiva hämnd för en svågers svek samt för sin systers död. Sigurds son dör före fadern och Atles söner dör före honom själv, i båda fallen initierat av kvinnors hämnd.20 Det är inte enbart på nordiskt område som far och son-motiv ägnas utrymme. Grundläggande i den fornfranska genren chanson de geste är ett homosocialt perspektiv, det vill säga att män definieras i förhållande till andra män. I genren finns dock en uttrycklig konflikt mellan generationerna. Om fadern finns i livet skildras han som ambivalent gentemot sina söner om de behöver hans hjälp. I exempelvis Chanson de Roland överlever fadersgenerationen och den yngre generationen dör. De flesta unga män dör dessutom barnlösa. Jämförelsevis sätts i den franska riddarromanen fadersrollen åt sidan och sönerna agerar som sjävständiga vuxna redan i unga år.21 I riddarromanen ges männen en identitet som grundar sig på mannen som subjekt och individ. När det gäller det sistnämnda förefaller det som om Sigurd i Völsunga sagas höviska del i 18 Se vidare Ney 2007, s. 80 ff. 19 Jfr Gurevitj, Aron, J., Den svårfångade individen. Självsyn hos fornnordiska hjältar och medeltidens lärde i Europa, övers. Carl G. Liungman, förord Thomas Lindkvist, Stockholm 1997, s. 99 f., Kärfve, Eva, Konsten att bli människa. Individ och myt i medeltidens riddarvärld, Stockholm / Stehag 1997, s. 22 ff. 20 ”Atlakviða”, i Eddukv ði, strof 39, s. 325 (”Kvädet om Atli”, strof 40, s. 269), ”Atlamál”, i Eddukv ði, strof 80, s. 344 (”Atlamål”, strof 74, s. 280). 21 Kay, Sarah, The Chanson de geste in the Age of Romance. Political Fictions, Oxford 1995, s. 103.

161  Far och son-motiv i nordisk hjältediktning och saga varje fall kan stämma överens med bilden av riddarromanens yngling, medan fadern Sigmund i förhållande till Sinfjötle kan jämföras med fadersdiskursen i chanson de geste. Den ambivalens hos fadern i förhållande till sönerna som karaktäriserar den genren kommer till uttryck i Sigmunds vargbett, men framför allt i förgiftningen av Sinfjötle. Sigurds identitet i den mytologiska delen grundläggs i förhållande till ett genealogiskt perspektiv där i första hand hans far och farfars egenskaper ger honom status och anseende. Den identiteten försvagas i den höviska delen till förmån för relationer till sekundära släktband. Referenser Beowulf, ed. George Jack, Oxford 1994 Beowulf, övers. Björn Collinder, Stockholm 1955 Beowulf and the Finnsburg fragment. A translation into modern English Prose by John R. Clark, London 1911 Breisch, Agneta (numera Ney), Frid och fredlöshet. Sociala band och utanförskap på Island under äldre medeltid, Ser. Acta Universitatis Upsaliensis, Studia Historica Upsaliensia 174, Uppsala 1994 Den poetiska Eddan, övers. Björn Collinder, 2. uppl., Stockholm 1964 Eddukv ði, ed. Gísli Sigurðsson, Reykjavík 1998 Fritzner, Johan, Ordbog over Det gamle norske Sprog 1-III, Kristiania 1886–1896 Gurevitj, Aron, J., Den svårfångade individen. Självsyn hos fornnordiska hjältar och medeltidens lärde i Europa, övers. Carl G. Liungman, förord Thomas Lindkvist, Stockholm 1997 Jónas Kristjánsson, Eddas and Sagas. Iceland’s Medieval Literature, transl. Peter Foote, Reykjavík 1988 Kay, Sarah, The Chanson de geste in the Age of Romance. Political Fictions, Oxford 1995 Kärfve, Eva, Konsten att bli människa. Individ och myt i medeltidens riddarvärld, Stock­ holm / Stehag 1997 Ney, Agneta, Drottningar och sköldmör. Gränsöverskridande kvinnor i medeltida myt och verklighet ca 400–1400, Möklinta 2007 Snorres Edda, övers. Karl G. Johansson och Mats Malm, Stockholm 1991 ”Völsunga saga”, i Fornaldar sögur Norðurlanda, band 1, ed. Guðni Jónsson, Reykjavík 1976 Völsungasagan, övers. Inge Knutsson. Inledning Staffan Bergsten, Lund 1991

Saga Reflections in Karen Blixen’s Texts (with Focus on Grjotgard Ålvesøn og Aud) Ieva Steponavičiūtė Vilnius University

Feeling quite an intruder in this honourable assembly of experts of the Viking age, I will start by quoting a person who stands closer to my own field of research, namely Jorge Luis Borges, one of the key figures in modern Western literature. In his book Old Germanic literature, he writes: Among ancient Germanic literatures, the Scandinavian is beyond comparison the richest and the most manifold. The first texts written in England and Germany have value because they announce, or we make us believe that they announce what would be written later: Milton’s glory casts light backwards on Cynewulf and the Song of the Nibelungs anticipates Richard Wagner. But the Old Norse literature possesses value in itself and those who study it can safely look away from Ibsen and Strindberg.1

Although a professor of literature, Borges was no real specialist in sagas and his book is written more from a fiction writer’s perspective. Yet most of us will probably agree with his idea of the autonomous value of the Old Norse heritage and vast possibilities it offers for researchers. As it does for authors of fiction: for hundreds of years, the classical Norse literature has served as an object of inspiration, imitation and transformation for Nordic writers. And if those interested in 1 Orig. De las antiguas literaturas germánicas la más compleja y rica es incomparablemente la escandinava. Lo que al principio se escribió en Inglaterra o en Alemania vale porque prefigura, o porque imaginamos que prefigura, lo que se escribiría después; la fama de Milton ayuda a la fama de Cynewulf y el Cantar de los Nibelungos anuncia a Wagner. En cambio, la antigua literatura nórdica vale por cuenta propia; quienes la estudian pueden prescindir de la evocación de Ibsen o de Strindberg. (Borges 1951: 76) Approaching the Viking Age. Proceedings of the International Conference on Old Norse Literature, Mythology, Culture, Social Life and Language. Edited by Ērika Sausverde and Ieva Steponavičiūtė. (Scandinavistica Vilnensis 2). Vilnius University Press, 2009. https://doi.org/10.15388/ScandinavisticaVilnensis.2009.2.10 Copyright © 2019 Authors. Published by Vilnius University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

164  Ieva Steponavičiūtė sagas might forget Ibsen and Strindberg, to study Ibsen or Strindberg without at least minimal knowledge of Old Norse literature is hardly possible, for one would risk missing something very important in their work. Both Strindberg and Ibsen,2 just like many Nordic authors before and after them, such as Adam Oehlenschläger, Selma Lagerlöf, Sigrid Undset, Johannes V. Jensen, Halldór Laxness, William Heinesen, Torgny Lindgren, Herbjørg Wassmo, Svava Jakobsdóttir, Einar Már Guðmundsson and lots of others, borrowed from sagas and Eddas motifs and stylistic devices and replayed them in their own artistic key. The Danish author Karen Blixen (1885–1962) is no exception in this respect. Allusions to “Gylfaginning” and “Völuspá”, paralleling, albeit often ironic, of female characters to saga women or valkyries (for example, Miss Malin in “The Deluge of Norderney” or Ehrengard in “Ehrengard” respectively), the intellectually cool narrator, showing little emotional involvement with the characters’ fates that she relates, insertion of verse into the narrative, expression of a character’s emotional state through changes in his / her physical appearance rather than explicit psychologising – all these aspects can be studied as reflections of Old Norse literature in Blixen’s authorship. Thus it is surprising, that despite these affinities and despite ample cases of Blixen’s play with Biblical and classical plots,3 there is only one story that is an explicit “remake” of a saga narrative. And this story is “Grjotgard Ålvesøn og Aud”. The text, which has never been fully completed, was written when Blixen was in her twenties. It was first published in 1962, shortly after the author’s death, based on a typewritten copy by Blixen’s brother Thomas.4 A somewhat different version was published in 1985 in Blixe­ niana, the publication of the Blixen society, based on a manuscript from



2 One can refer here to The Folkunga Saga by Strindberg, or The Vikings at Helgeland by Ibsen, to mention just a few examples. 3 Consider, for example, transformations of “The Dead” by Joyce in “Babette’s Feast”, “Ball-of-Fat” by Maupassant in “The Heroine”, “The two Baronesses” by H.C. Andersen in “A Country Tale”, “The Seducers’s Diary” by Søren Kierkegaard in “Ehrengard” or the naughty recount of the escape of the Holly Family to Egypt in “The Deluge at Norderney”. 4 Blixen 1962.

165  Saga Reflections in Karen Blixen’s Texts Blixen’s own hand.5 I will refer to the latter version in my analysis, since the former one, as the editors of Blixeniana claim, might contain Thomas Dinesen’s own contribution (Lasson 1985: 13). It is possible, however, that some differences in the two versions are due to Blixen’s own rework of the story, if she, for example, considered its publication later in her career.6 The Blixeniana version does exhibit traits of Blixen’s mature writing: elaborate syntax and also so typical of Blixen semantic ambiguity resulting in difficulty for the reader to (re)construct the fabula behind the sjužet, i. e. to answer the question what “really” happened in the story. Yet another typical trait would be the art of storytelling as the ultimate topic of the text and a sense of narrative pleasure encoded therein. The plot of “Grjotgard Ålvesøn og Aud” is based on a series of episodes from “Saint Óláf ’s Saga” which is part of Heimskringla, the history of Norwegian kings written, as most sources agree, by Snorri Sturluson. These episodes relate the feud between the landlords of Egg and King Óláf which indirectly leads to the climax of the saga – Óláf ’s death at the Battle of Stiklarstath.7 This storyline starts around the middle of the saga (Chapter 106) when we are told about Olvir of Egg, the leader of farmers who kept sacrificing to heathen gods, despite the king’s effort to convert them to Christianity. Óláf had him captured and executed, as this was his way to spread the new faith in Norway. Olvir’s good looking and rich widow Sigrith was then married by Óláf to one of his men – Kálf Árnason, who later turned away from the king and even took part in his killing.8 As the text implies, Kálf did it on Sigrith’s demand to avenge the killing of the two sons – Thórir and Grjótgarth – she had had with Olvir (Chapter 183). Thórir’s death is recounted in one of the most artistic episodes of Snorri’s saga, containing a dramatic dialogue in which the king and

5 Blixen 1985. 6 Unfortunately, the editors of Blixeniana do not provide any information about the possible dating of the manuscript that they have published. 7 The English transcription of Old Norse names and place-names is taken from Lee M. Hollander’s translation, see Snorri 1964. 8 Snorri’s text does not state explicitly whether it was Kálf Árnason or his namesake who stroke the fatal blow, but Kálf Árnason’s participation in the battle of Stiklarstath is beyond doubt (see Chapters 223–225 and 229).

166  Ieva Steponavičiūtė Thórir try to outwit each other (Chapter 165). King Óláf is staying with his men at Thórir’s farm. Thórir is now eighteen, a handsome strong man and a wealthy landlord. He gives Óláf a royal treat and everyone is happy till Óláf talks to one of his men – Dag, whom he believes to possess clairvoyant powers. Dag claims that Thórir has received payment from the Danish King Knút for promising to kill Óláf. The proof of that, Dag says, is the golden ring that Thórir is wearing on his right arm. Thórir is unmasked under the dialogue with the king – a brilliant example of Snorri’s understatement technique: “How old a man are you, Thórir?”, asked the king. “I am eighteen years old”, he replied. The king said, “A big man you are for your age, Thórir, and a fine fellow.” Then the king put his hand around Thórir’s right arm and stroke it above the elbow. Thórir said, “Gently, sire! I have a boil on my arm.” The king held on to his arm and felt something hard underneath. The king said, “Haven’t you heard that I am a healer? Let me see that boil.” Then Thórir saw that it would not do to conceal it any longer and took off the ring and showed it to the king. The king asked whether it was a gift from King Knút. Thórir said he would not deny it. (Snorri 1964: 455–456)

Thórir is executed the next day and the next chapter briefly tells of his brother Grjótgarth who sets out to avenge Thórir’s death by attacking the king’s property and people, but is killed one night in a fight with Óláf ’s men. These are the events that constitute the main frame of reference for Blixen’s text and are incorporated in it without considerable alteration. Blixen does not change the outcome of the events, nor does she distort the family relations. Blixen’s Tore (= Thórir) and Grjotgard (= Grjótgarth) are also brothers, whose father Ålve (= Olvir) has been executed by King Olav (= Óláf).9 Tore also receives the king in his home in Blixen’s text. Here too, the king has Tore killed after having

9 The names are spelled here as they appear in the Danish text by Blixen.

167  Saga Reflections in Karen Blixen’s Texts talked to Dag – an act that provokes Grjotgard’s revenge. So in general, Blixen’s story seems to respect the rules of good reading that Umberto Eco (2004) describes in his article “On Some Functions of Literature” in which he further develops his concept of the limits of interpretation. Eco speaks of the paradox of great literature: although fictive, it is in possession of truth that is much more difficult to deny than it is in the case of historical facts. It is possible that one day a proof is found that will change our knowledge of the circumstances of Napoleon’s death, but Sherlock Holmes will forever remain a bachelor, as it is absolutely true that Hamlet never marries Ophelia, and who can respect those who claim otherwise? A good interpreter, Eco insists, won’t dare to violate the truth of “great narratives” in which people have invested their emotions and which make us understand that things don’t develop the way we want, but follow their own, albeit sometimes tragic, course: “The function of unchangeable stories is precisely this: against all our desires to change destiny, they make tangible the impossibility of changing it” (Eco 2004: 14–15). With the original story line left unaltered, Blixen’s text demonstrates awareness of this basic function of literature, as well as respect for the narrative in which Nordic people of many generations have made their emotional investments. Yet, I wouldn’t bother you here with an analysis of a text if it was nothing more than a respectful repetition of the “factual” truth of a story already known. “Grjotgard Ålvesøn og Aud” deserves the reader’s attention, because Blixen imagines between the lines and episodes of the saga, slightly alters some of its details and creates her own story with its own dramas and its own ideology. The very first sentence of Blixen’s text, which contains an obvious allusion to the saga written by Snorri, announces a new perspective: ‘A tale written down by a chieftain is also being told by a slave.’10 In contrast to her precursor, Blixen introduces into the story an intradiegetic first person narrator. It is Grjotgard’s slave Finn, who through a series of flashbacks (a method seldom, if ever, encountered in saga literature) relates the events surrounding Tore’s death. Part of his 10 Orig. Et Sagn, som er nedskrevet af en Høvding, fort lles ogsaa af en Tr l. (p. 52)

168  Ieva Steponavičiūtė recount, namely the king’s stay at Tore’s place (including the dialogue), is a direct borrowing from the saga, yet Finn also tells of events which might have preceded and followed this episode and which are not mentioned in Snorri’s text. The explicit narrator’s presence in the text has an ambiguous effect. On the one hand, it creates the illusion that we are going to hear a more truthful story than that presented by its precursor. Finn tells of events that he has witnessed, while Snorri, as we know, writes his story down more than 100 years after the events took place. On the other hand, Finn is a story-teller who tells fairytales to children, and who, as we are told, ‘creates where he forgets.’11 So the figure of this unreliable narrator telling a fragment of Norwegian history seems to be a metafictional suggestion that the true representation of history is hardly possible, since we cannot approach history other than through someone else’s narrative.12 Another change on the level of the narrative structure in “Grjotgard Ålvesøn and Aud” is the rearrangement of dramatis personae. In the saga, Grjótgarth is presented very briefly in an episode developed with much greater narrative economy and less stylistic finesse than the preceding one about Thórir. In Blixen’s text, he becomes a title character on whose fate the story is focused. First of all, Blixen supplies her Grjotgard with a CV. We are told that he has been away on a Viking raid for two years, doing nothing at the moment (a fact that annoys both Finn and Grjotgard’s households), but is secretly planning to go to Denmark to see King Knut, King Olav’s greatest enemy (p. 58). Grjotgard is also granted an existential condition: situated somewhere between two cultures, the heathen and the Christian, he belongs to neither of them. He tarries revenging his father’s death, but cannot forgive it either. And then finally, Blixen sets this character in an emotional relationship to other characters, first of all to his brother Tore. With the help of numerous dialogues that Grjotgard enters in Finn’s recount, we find out that he is bearing a grudge against Tore. One of the reasons for this rift is 11 Orig. havde han glemt noget, digtede han det. (p. 55) 12 This implication is reinforced by the incorporation into Finn’s story of yet other narratives – that of Groa, the female slave in Aud and Tore’s house, and that of the messenger, both of whom witness the events taking place in Tore’s house where Finn himself was not present.

169  Saga Reflections in Karen Blixen’s Texts Tore’s becoming friends with their father’s killer. This can be concluded from Grjotgard’s answer to Finn’s question why he is not at his brother’s farm when the king is there: “I am not there for two reasons,” “First, because I am no friends with King Olav. I was twelve when he had my father killed at Egge; he is not likely to think that I’ve forgotten that as well as Tore. The king doesn’t know either if I am just as good a Christian as my brother.”13

Of the second reason, Grjotgard tells almost nothing (‘The second reason is that I don’t like being at my brother Tore’s place’14), but Finn’s words help us to understand that it is jealousy: ‘You don’t like to be there in the morning, my lord, and you don’t like it much to be there at midday, but least of all you like it, when the bedtime approaches .’15 While Grjotgard was away on a Viking expedition, Tore, as we find out, married Aud, a girl whom Grjotgard had loved. Thus Blixen creates in her text an additional semantic field of the two brothers’ discord, introducing a love triangle as one of its elements. Aud does not exist in “Saint Óláf ’s Saga” (we are only told that Thórir was married, and that the marriage made him rich).16 Blixen not only invents this character, but also makes it the key agent in the development of the narrated events. As we already know, Aud is one of the reasons for the two brothers’ alienation. It is also her who incites Grjotgard’s revenge when he comes to Tore’s place after having received Aud’s message of his death:

13 Orig. „Jeg er ikke med af to Grunde” „Først fordi jeg ikke staar mig godt med Kong Olav. Jeg var tolv Aar, da han lod min Fader dr be paa Egge; han tror nok ikke, jeg har glemt det saa godt som Tore. Saa ved Kongen hellerikke, om jeg er saa god Kristen som min Broder.” (p. 56) 14 Orig. „Men den anden Grund er det, at jeg ikke synes godt om at v re hos min Broder Tore.” (p. 57) 15 Orig. „Du synes ikke saa godt om at v re der om Morgenen, Herre, og du synes ikke meget godt om det om Middagen, men mindst synes du dog om det, naar det gaar mod Sengetid .” (p. 57) 16 Snorri 1964: 454.

170  Ieva Steponavičiūtė “in his own farm he was killed, in his own house! I have sat here and heard them shout for revenge, your kin and mine. Thank you, Grjotgard, for I know now, I will get my revenge.”17

The active role Aud plays in the story is not, however, limited to the two things just mentioned. There are implications in the text that she, paradoxically enough, plays a role not only in the events invented by Blixen, but also in those that are borrowed directly from the saga, i. e. Tore’s unmasking and death. The idea that Aud was the actual reason for Tore’s death is first proposed by Grjotgard. This happens at the very end of the story, when Grjotgard and Aud finally meet again besides Tore’s body and exchange obscure mutual accusations: Grjotgard said: “It’s not sure, if you have to thank me, Aud. Would you like to know what I was thinking while riding here? It was that you are guilty of Tore’s death more than anyone else. Hadn’t Tore had you in his house, he would have never betrayed the king.” “I did not advise him”, said Aud. “Yes, you did, Aud,” said Grjotgard, “I had known him before you knew him; I know there was somebody behind him, otherwise he wouldn’t have done that. Were you a man, I would kill you.”18

I will shortly come back to the interpretation of Grjotgard’s words and of Aud’s role in Tore’s death, but let me first quote Aud’s reply:

17 Orig. „i sin egen Gaard er han blevet dr bt, i sit eget Hus! Jeg har siddet her og hørt dem raabe om H vn, din Æt og min. Aa, du skal have Tak, Grjotgard, for, at jeg ved nu, jeg faar H vn.” (p. 69) 18 Orig. Grjotgard sagde: „Det er ikke saa sikkert, at du skal takke mig, Aud. Vil du vide, hvad jeg t nkte, mens jeg red herover? Det var, at mest af alle er du Skyld i Tores Død.” „Havde Tore ikke haft dig i sit Hus, saa havde han aldrig sveget Kongen.” „Jeg raadte ham ikke,” sagde Aud. „Jo, du raadte ham, Aud”, sagde Grjotgard, „jeg har kendt Tore, inden du kendte ham; jeg ved det nok, at der har v ret nogen bag ham, eller havde han ikke gjort det. Hvis du var en Mand, saa vilde jeg dr be dig.” (p. 57)

171  Saga Reflections in Karen Blixen’s Texts “you let me hear that I have caused my husband’s killing; now I will tell you, that you are guilty in your brother’s death.” “It’s true, Grjotgard I know how he lay awake at night, when he had heard about your reputation in daytime. He never had another thought, I know it best.”19 The dialogue ends the story but, as it typically happens in Blixen’s

writing, the end does not bring it to closure. Although imagining what is left unsaid in the saga, Blixen creates her own blanks, gaps and indeterminacies, provoking the reader to go on imagining. Grjotgard and Aud talk in riddles: it is obvious that they blame each other for Tore’s death, yet it remains unclear how exactly they think the other could have brought it about. It is also strange that Grjotgard accuses Tore of betraying the king (Olav?), when earlier in the text he blames his brother for forgetting their father’s death and hosting the killer. And yet it seems possible to solve these textual puzzles and make sense of the dialogue as well as the whole story. In the present paper, this will be done by interpreting different hints scattered throughout the story and fitting them into each other, as well as by analysing the allusions and references of the text to the saga. One can start by saying, that the king whom Grjotgard refers to does not have to be Olav, the king mentioned most extensively in the text. Having in mind his earlier grudge against Tore for forgiving Olav their father’s death, it is more reasonable to think that Grjotgard has in mind King Knut (having heard Aud’s messenger, he knows now that Tore had sworn allegiance to the Danish king, just like Grjotgard was planning to do himself). Thus he seems to be blaming Aud for convincing her husband to break this allegiance and join Olav, in this way delivering him right into the hands of their family’s enemy (‘Hadn’t Tore had you in his house, he would have never betrayed the king’). Aud, in her own turn, does not deny her own involvement in Tore’s death, what she does not accept is the accusation of exercising influence 19 Orig. „du har nu ladet mig høre, at jeg har raadt min Mands Bane; nu vil jeg sige dig, at du er Skyld i din Broders Død.” „Sandt er det dog, Grjotgard,  jeg ved nok, hvordan han laa vaagen om N tterne, naar han havde hørt dit Ry om Dagen. Han havde aldrig nogen anden Tanke, det ved jeg bedst.” (p. 57)

172  Ieva Steponavičiūtė on her husband (‘I did not advise him’). At the same time, she claims that the initial reason for Tore’s death lies with Grjotgard himself. She indirectly accuses him of taking no action against Olav, at the same time implying that Tore was planning to kill Olav. It is possible to claim so, because earlier in the text we are told by Finn that people regret Grjotgard’s passivity and hope that he is planning to join the King of Denmark (pp. 57–58). So this must be the reputation that according to Aud did not let the younger brother Tore sleep at night and forced him to commit the revenge himself. That Blixen’s Tore did not forget his loyalty to King Knut and was indeed planning to avenge his father’s death is of little doubt, as Blixen changes one small detail in the scene borrowed from the saga: Tore retains the golden ring on his arm (a sign of loyalty), whereas he takes it off in the original text.20 It seems, however, that Aud was resisting this scenario, according to which the revenge should be executed by Tore. As we shall see, she had a different plan for which it was even necessary to have Tore killed. There are several things that suggest Aud’s involvement in Tore’s death: not only Aud’s hawkish eyes (p. 59), as a symbolic expression of her untamed and dangerous nature, but also her behaviour during the episode of the king’s visit. Aud is described as absolutely undisturbed. Despite warnings about the change in the king’s mood, she tells her husband himself to serve the king (p. 65), so, we may guess, he can discover the ring. Another support for the proposition that Aud may have plotted Tore’s death can be found in the character of intertextual relations of the episode to its hypotext,21 i. e. the saga. It follows the narration in the hypotext very closely, tending towards direct quotation, thus we can look for suggestions for its interpretation in the saga. In Snorri’s text, there is a great deal of obscurity surrounding the circumstances of Thórir’s unmasking. The king is told about Thórir’s complot by Dag. Dag informs the king about the ring, which, Dag claims, Thórir does not let anyone see. Shall we believe that Dag knows of the ring because he possesses clairvoyant powers, the way the king believes it? The narrator of the saga seems, however, to be ironic (or at least doubtful) about Dag’s 20 Cf. Blixen 1985: 65 and Snorri 1964: 456. 21 An anterior text, which a given text (“the hypertext”) transforms. (Genette 1982: 5)

173  Saga Reflections in Karen Blixen’s Texts skills and does not share Óláf ’s trust in him: the only proof he provides for Dag’s miraculous talent is when he in the earlier episode relates how Dag finds the hiding place for property that he himself was accused of stealing (Snorri: 453–454). So a more reasonable explanation for Dag’s prophecy would be that he knew about the ring from somebody very close to Thórir, and why could it not be his wife? There is one more direct intertextual allusion to the saga in Blixen’s text (this time relating directly to Aud’s character) that should be mentioned in this argument. By inciting revenge for Tore’s death, Aud partly takes over the role performed in the saga by Sigrith, Thórir and Grjótgarth’s mother. Sigrith demands from her husband Kálf to avenge the death of her first husband and their sons and openly rejoices when Kálf ’s brother is killed, for now she can expect Kálf to take action: “It is good that you had to bear that from the king, because it is likely that him you will wish to avenge, even though you do not care to avenge the wrongs done to me” (Snorri: 478). Although this parallel does not point directly to Aud’s involvement in Tore’s death, it suggests that she is at least happy with it. Aud has a similar goal as Sigrith (to incite Grjotgard to act), but her motives are different, and so is her function in the text. While Sigrith can be said to play the role of the guardian of the honour code that was gradually disintegrating with the advance of Christianity described in the saga, Aud is concerned with something other than the restoration of family honour. She definitely was not planning her husband’s death in order to marry Grjotgard: a scenario that would fit a cheap melodrama, but not a text by Blixen with her love of paradox. The key to Aud’s motives are to be found in Aud’s words to Grjotgard before her marriage with Tore: ‘If I were you, I would never let any other man be equal to me.’22 Just as Sigrith, Aud desires that Grjotgard should take action, but the revenge she calls for seems to be only a means for her to make Grjotgard, the man she loves, a hero. Aud could not accept Grjotgard’s passivity, but as Tore’s wife she could not incite Grjotgard’s action against Olav. Now with Tore dead, she has a right to demand 22 Orig. „Hvis jeg var dig, saa vilde jeg ikke lade nogen anden Mand v re lige med mig.” (p. 60)

174  Ieva Steponavičiūtė revenge for her husband and gives Grjotgard a chance to face an opponent worthy of him. Aud’s aspirations, compared to those of Sigrith’s, have a somewhat romantically demonic flavour, though her very actions, both explicit and implied, do not contradict the behaviour of saga women. We know of saga women, who could consciously contribute to their husband’s death, had they a “good” reason for that (the classical example would be Hallgerd from Njal’s Saga who spares her hair that could have saved Gunnar’s life as she cannot forget the slap on her face, she has once received from him). This seems to be the message that Aud’s character appears to convey: women want to love men who deserve their love, and if men cannot find one themselves women sometimes try to arrange for them a possibility to prove their brilliance. But in order to interpret the message of the whole story, we, of course, have to take into account its narrative situation. It is the explicit narrator Finn who arranges the events into a narrative, so it is important to know what he thinks about the whole affair. There is little doubt that Finn regrets that everything did not go the way Aud had planned. Also he had had great plans about Grjotgard’s future, and he is sad to admit to his listeners that his predictions about Grjotgard’s meeting with Olav did not come true: “ Sometimes, when many people listen to you, and there’s great news in the air, one cannot help seeing great pictures and predicting in a great many words. While one speaks, one believes in what one says, but later finds out, that it was nothing.”23

Finn does not tell us how it went with Grjotgard’s revenge, but he does not need to, for we know it from the saga the factual truth of which Blixen’s text respects. Grjótgarth, as Snorri records, suffers a pathetic death: surrounded by Óláf ’s men, he rushes out uttering brave words, 23 Orig. „ Men sommetider, naar mange Folk hører paa én og der er store Tidender i Luften, kan man ikke lade v re at se store Billeder for sig og spaa i mange Ord. Mens man taler, tror man paa det selv, men bagefter ved man nok, det er ingenting.” (p. 67)

175  Saga Reflections in Karen Blixen’s Texts but in the darkness directs his sword at the wrong man and is slain at once (Snorri 1964: 457). Blixen’s text allows us to guess what kind of future Finn had imagined for his master instead of this one. At one point of his story, Finn reminds Grjotgard of an episode, when he, being only fifteen, killed the bear which Kalv (= Kálf) had been chasing for days (p. 57–8). Thus he implies that he was greater than the man whom we know as one of the key figures in the scene of the battle of Stiklarstath.24 It is true, however, that there is a reservation in the saga that Snorri, the historian, makes about Kálf ‘s role in the slaying of Óláf, saying that “Men disagree as to which Kálf [Kálf Árnason or Kálf Árnfinnsson] wounded the King” (Snorri 1964: 515). Yet, Snorri, the story teller, seems to be quite sure that it was Kálf Árnason, since in the very same episode he quotes a verse of skaldic poetry glorifying him. Finn, as we remember, is also a storyteller. But it is a strange story he tells, a story that lacks cohesion. Finn starts a tale about Grjotgard and then regresses to fairy tales and gets angry when his listeners ask him to come back to the tale (p. 55). Actually he never brings it to the very end, to Grjotgard’s death. With all his hopes about Grjotgard’s glorious meeting having collapsed, he seems simply to be lacking material for his art. He is deprived of a possibility to create the story he wishes for and, we may now assume, this is the true reason why he is left to grieve over his master’s death forever.25 Blixen’s story was written at the very start of her career, but its message is so typically Blixenish: life has value only when it can be transformed into art and allowed to enter the common cultural memory. We will later find similar ideas in Out of Africa, or “The Sorrow Acre”, for example. The message itself may sound trivial or too high-flown for the modern ear, yet the irony of Blixen’s texts is that they do not state their message explicitly, but often trick the reader into arriving at simplified conclusions. The very way it is done – through creation of 24 Also Aud presents Grjotgard with a plan that could have led him directly to the glory at Stiklarstath, had he listened to it. She urges him to seek allies and assemble an army, instead of attacking Olav immediately – the way he wants it and supposedly does, as the episode of his death in the saga suggests. (p. 68) 25 Orig. Grjotgards Tr l Finn, som aldrig ophørte at sørge over sin Herre . (p. 52)

176  Ieva Steponavičiūtė a semantically dense and opaque (inter)textual field – can hardly be called trivial and simple. I will conclude my analysis here by saying that what I just did was to look at the two texts from the traditional intertextual perspective, using the hypotext, i. e. the saga as an interpretative key to open up the semantic contents of the hypertext, i. e. Blixen’s story. Often it is also interesting to apply the reversed, the so-called Borgesian model of intertextuality, and speak of the later text as a key to the earlier one.26 Though I am afraid, this could be considered too far fetched, since “Grjotgard Ålveson and Aud” is not a very well known story and Blixen was no great writer when she composed it. And yet, if you read the episodes about Thórir and Grjótgarth in the saga after having read Blixen’s text you cannot help concentrating on the things that allowed the engendering of Blixen’s story: on the blank spaces of the saga, on things that raise suspicion or somehow do not make sense. On questions like: how is it possible that Grjótgarth was not with his brother if the latter was plotting something, especially when the saga says that he was around in the area, and was Thórir really guilty of the crime he was killed for, or who might have told Dag about the ring if he was not clairvoyant? And although it is nothing revolutionary to say that “Saint Óláf ’s Saga” is more a piece of art, than history, the reading of Blixen’s story raises this awareness even more, since you feel that Snorri’s text attracts you by its power to create suspense by what is left unsaid, or is said ambiguously, thus making space for reader’s own imagination. Actually professor Vésteinn Ólason in his enlightening presentation yesterday and in his book Dialogues with the Viking Age27 devoted to the narrative specifics of the sagas of Icelanders illustrated the idea, that modern texts lend their reading strategies for reading sagas. They do encourage to study what is implied or is decontructive alongside with what is explicit and logically sound. But what is even more important, is that such texts like the one I have just discussed simply make you read sagas. I must confess that 26 Cf. his famous phrase: “Every writer creates his own precursors. His work modifies our conception of the past, as it will modify the future.” (Borges 1964: 199–201) 27 Vésteinn Ólason 1998.

177  Saga Reflections in Karen Blixen’s Texts my copy of Heimskringla was very dusty when I found it, but I am happy that I did it, and I did it because of Blixen’s text. But this is only part of the story. I turned to the latter because I knew that our Center of Scandinavian Studies would host a conference on Old Norse culture. So after all, maybe one shouldn’t be too arrogant about Strindberg and Ibsen (and one can hardly fail noticing the irony in Borges’ words), or Blixen for that matter, because we have here a cycle of mutual dependence: we read sagas to better understand modern literature, but it is modern literature that brings us back to sagas and to a very high extent secures them their sustainable position within the national or even transnational cultural memory. This is no new thing to say (these processes have been dwelt upon by T. S. Eliot, Jorge Luis Borges and Harold Bloom among others), and as our reading experience shows, it still holds. Bibliography Blixen, Karen. “Grjotgard Ålvesøn og Aud”, in: Osceola. Ed. by Clara Svendsen. København: Gyldendals Julebog, 1962, pp. 10–37. Blixen, Karen. “Grjotgard Ålvesøn og Aud”, in: Blixeniana 1985. Ed. Hans Andersen and Frans Lasson. København: Karen Blixen Selskabet, 1985, pp. 52–69. Borges, Jorge Luis (in collaboration with Delia Ingenieros). Antiguas literaturas germánicas. México: Fondo de Cultura Económica, 1951. Borges, Jorge Luis. “Kafka and his Precursors”, in: Labyrinths. New York: New Directions, 1964, pp. 199–201. Eco, Umberto. “On some functions of literature”, in: On Literature. Trans. Martin McLaughlin. New York City: Harcourt, 2004, pp. 1–15. Genette, Gérard. Palimpsests: Literature in the Second Degree. Ed. and trans. Channa Newman and Claude Doubinsky. Lincoln: University of Nebraska Press, 1982. Lasson, Frans. “Om Karen Blixens ungdomsfortællinger”, in: Blixeniana 1985, pp. 11–14. Snorri Sturluson. Heimskringla. History of the Kings of Norway. Trans. Lee M. Hollan­ der. Austin: University of Texas Press, 1964. Vésteinn Ólason. Dialogues with the Viking Age. Narration and Representation in the Sagas of the Icelanders. Reykjavík: Heimskringla, 1998.

Scenes from Daily Life in the Sagas of Icelanders: Narrative Function Vésteinn Ólason The Árni Magnússon Institute for Icelandic Studies, University of Iceland

This topic may sound trivial, and so it is, in a sense. It should be emphasized, however, that this paper does not deal with scenes from daily life in the sagas from the point of view of folklife studies – although that might in itself be interesting – but strictly from a literary angle. It is – or was – frequently assumed that the saga was a rather simple and pure narrative form – one of the “Einfache Formen” (to quote an old book title) – like the folktale and the heroic song.1 The saga is in some ways similar to the simple forms: action is all-important, but the saga’s development as a written form has opened up possibilities for the occasional inclusion of elements that enlarge and add new dimensions to the picture drawn by the narrative, thus raising the question of how this might affect the potential that the genre has for expression of ideas and emotions. In Dialogues with the Viking Age, a book that I published in 1998, I attempted to describe the sagas as a kind of literature, outlining their formal characteristics as narratives as well as their relationship to a particular society, a particular culture and its memories. There I maintained that the people who wrote the sagas tried to find their bearings in a turbulent world by conducting a kind of dialogue with their own past by 1 See Jolles 1968. Jolles discusses the Icelandic sagas briefly in the context of Sage, pp. 66–75. Although a firm believer in oral sagas, Jolles does not classify them as ‘Einfache Formen’; they are, he says, “an und für sich ebensowenig Einfache Form, wie die Viten, die in den Acta Sanctorum gesammelt wurden. Auch hier haben wir, was wir Vergegenwärtigung einer Einfachen Form oder aktuelle Form genannt haben. Aber darüber hinaus ist auch die gefestigte mündliche Überlieferung, die in den Handschriften schriftlich fixiert wurde, noch keine Einfache Form… auch sie ist gegenwärtig und damit in gewissem Sinne schon Kunstform” (p. 71). Approaching the Viking Age. Proceedings of the International Conference on Old Norse Literature, Mythology, Culture, Social Life and Language. Edited by Ērika Sausverde and Ieva Steponavičiūtė. (Scandinavistica Vilnensis 2). Vilnius University Press, 2009. https://doi.org/10.15388/ScandinavisticaVilnensis.2009.2.11 Copyright © 2019 Authors. Published by Vilnius University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

180  Vésteinn Ólason telling stories about important and interesting events in the lives of their ancestors. The saga writers were deeply anxious about the changes taking place in their society that presented serious challenges to their traditional values. Christianity offered them a strong world view that answered all questions about matters of morality as well as public life. That world view had never fully gained ground in secular affairs, although this was now occurring as the Norwegian king sought to extend his influence and authority in Iceland. However, Icelanders knew tales and poems from the past that described events, often tragic events, in the lives of people who were considered admirable, although these people did not necessarily behave admirably from the point of view of the Christian morality. The duty to take revenge for certain offences against their families led them into feuds that often had tragic consequences. Therefore, the Icelanders of the thirteenth century looked nostalgically to a world that was basically tragic, but in which people had the choice to act with dignity at the risk of losing their lives or those who were dearest to them. The thirteenth-century world seemed to offer simpler solutions, but something was missing. The sagas react in different ways to this situation, and there is, in my opinion, a development in the reactions. What is under discussion here may be seen as a side issue in the analysis of the saga genre, but it is usually helpful to approach complex problems from many sides. A study of one particular kind of literature in a historical context inevitably makes one aware of the interplay of genres, or discourses if you like, that constitute the background of a particular genre and define it. Part of my previously mentioned project was therefore to compare the sagas with other kinds of narrative that could have formed this background, but also to modern narrative. These remarks about daily life in the sagas are to be seen in this context.2 When the sagas started to be known outside Iceland, they were seen primarily as interesting expressions of the spirit of a primitive Germanic or Northern society with a peculiarly developed sense of honour and a fine narrative tradition. Accidentally, many of their characteristics found

2 Another extension or continuation of the argument in Dialogues with the Viking Age is to be found in a recent article by the present author, 2007a. See also: Vésteinn Ólason 2007b.

181  Scenes from Daily Life in the Sagas of Icelanders an echo in the literary taste of the nineteenth century, a taste that was formed by the realistic tradition in novel writing. For all the differences between saga and novel, the characterization and restrained style of the sagas was much closer to the nineteenth-century tastes than the style of medieval narrative forms already familiar to people, such as courtly romances or saints’ lives. Moreover, the sagas told stories of ordinary people rather than limiting themselves to kings and aristocrats, and this harmonized well with the tastes of new groups of readers. Several scholars declared that there was an unmistakable relationship between the sagas and the novel.3 Historically, it is easy to prove that such affinities as there are between the sagas and the novel must be accidental; there is no direct link. When the modern novel arose in totally different social circumstances, the sagas were totally unknown outside Iceland (the first important novelist who knew anything about the sagas was Sir Walter Scott (Wawn 2000: 66).4 Exactly for this reason, however, it is interesting to study the narrative form of the sagas and ask what it is that has made people compare them with the novel. Scenes from daily life can be interesting from this point of view, because no genre is as rich in such scenes as the novel. The function of such scenes affects the relationship between content and form, and demonstrates that when form is filled with a new kind of content, it will be changed. Experience shows that when scholars are dealing with groups of texts they often tend to work with simplifications, plot summaries in the case of narrative literature. The scholar looks at one important link after another in the chain of narrative, investigates what is added to what happened previously and what possibilities are opened for further development of the plot. This is practical, but it means that many scenes and episodes, not to mention smaller segments of the text, that seem less important escape attention, although such elements may be revealing and interesting upon a closer look. Many scenes from daily life are of this sort. Although structurally unimportant, they add nuances to the texture of the works that ought not to be overlooked in the analysis and interpretation of these works.

3 See, for instance Ker 1957: 183. 4 See also Simpson 1973.

182  Vésteinn Ólason In Íslendingasögur, interest is directed only or almost exclusively to memorable events leading to, advancing and resolving conflicts concerning honour.5 Describing daily life for its own sake is definitely not one of the sagas’ concerns. Nevertheless, we can find scenes from daily life in the sagas, and the study of such scenes can throw interesting light on their nature as narratives. The Icelandic saga gradually developed into a separate kind of written narrative during the second half of the twelfth and the first half of the thirteenth century. The saga has many roots: on the one hand in the written literature of the Middle Ages, historical works (historiae), chronicles, romances and saints’ lives, but also in oral narratives such as the heroic lay and the heroic tale, as well as historical or local legends of various kinds. Although the saga developed its own distinct characteristics, it did not sever all its ties to these different kinds of narrative; on the contrary, it developed in constant interaction with the types of texts (written or oral) surrounding it. Inclusion of scenes from daily life may widen the scope of saga narrative and change its nature in the process, and it may offer clues to its interpretation, to our understanding of the fate of the characters not otherwise easily decipherable. It is well known that many types of traditional narrative are indeed ‘closed’ texts, that is, they move through conventional steps that lead to a predictable end, and their world is composed of a finite set of elements that can be arranged in different ways but on the basis of fixed or nearly fixed rules, that can be seen as a kind of grammar. These rules govern the understanding as well as the creation of a tale: the tale refers directly only to its own kind. The best known example of this finiteness is the fairy tale or Märchen, as analysed by Vladimir Propp (1968). The heroic tale is a much looser concept, and it refers to a more varied group of texts than the fairy tale. Nevertheless, there have been many attempts to describe a basic form of tales about heroes using methods similar to those employed by Propp and his structuralist followers.6

5 A thorough treatment of the subject is found in Meulengracht Sørensen 1993: 187–248. 6 A survey of hero pattern studies is found in Taylor 1964 Best known of such studies is probably Lord Raglan‘s study, The Hero (1936); see also de Vries 1959: 194–208.

183  Scenes from Daily Life in the Sagas of Icelanders It is a fundamental feature of both the fairy tale and the heroic tale – or the heroic song – that practically every element in the text serves the action. This is not always obvious when we are looking at descriptive or introductory elements, but in fact they serve either to characterize a type of person filling a certain role, which is necessary for the action, or, perhaps, a type of environment, a conventional setting for such action as will occur. There is, of course, some room for variation in all of these elements, depending upon the convention. We must keep in mind that the world and action of written narrative can be just as conventional or closed as that of oral narrative, although the written form usually allows for more variation. Íslendingasögur have many conventional elements. They have hero types and typical patterns of action,7 but quite often there are elements in the texts that are not easy to classify as belonging to a fixed or finite set of saga-elements. Occasionally, we find scenes from daily life that seem to open a window to the world which surrounds the closed universe of the conventional saga. If such openings are found only exceptionally, the basic structure may remain intact, but if they become more numerous, if one peephole is replaced by many windows, we see a qualitative change in kind, from a closed to an open narrative. This change is most important, because it opens the narrative for new and various interpretations. Such a change towards a more open form was occurring in Íslendingasögur, causing them to make a relatively modern impression compared with many other literary works of the Middle Ages. As an illustration, five saga-scenes from daily life shall be discussed here: two describe meals, one describes hair-washing, and two show a man and a woman in the privacy of the bedchamber. There is hardly a more common act in people’s lives than having a meal. How it is done varies greatly depending on time and place and the relationship of the individuals involved, as we all know. In many modern novels we find detailed descriptions of such scenes. They are a convenient frame for dialogue and help to characterize the cultural environment of a place or a period and add local colour. Moreover, for many modern people writing and reading about a meal is a source of

7 These are discussed, for instance, by Lönnroth 1976: 61–82.

184  Vésteinn Ólason pleasure in itself. Such scenes are rare in Íslendingasögur, and those we find are certainly not included for the purpose of describing the enjoyment of a good meal. A gathering for eating and drinking, a banquet in a hall, is a conventional motif in Germanic heroic poetry as we know from Beowulf, the eddic lays about Atli, and from sagas. These banquets are very formal occasions, with emphasis on how people are seated according to status, and quite often a dialogue occurs which is crucial for the action to come. This motif is prominent in Icelandic sagas, such as Egils saga, Laxdœla saga and Njáls saga. In two sagas, however, we read about a different kind of meal, informal and much closer to daily life than a banquet. In Fóstbrœðra saga (“The Saga of the Sworn Brothers”) one of the two heroes, Þorgeirr Hávarsson, has sought quarters with a farmer during his wanderings around a sparsely populated area of Iceland. The farmer already has a visitor, a tramp called Butraldi, whom the saga describes as: “einhleypingr, mikill maðr vexti, rammr at afli, ljótr í ásjónu, harðfengr í skaplyndi, vígamaðr mikill, nasbráðr ok heiptúðigr” (ÍF, vol. 6, pp. 142–43). [a loner of no fixed abode. He was a large, powerfully built man with an ugly face, quick-tempered and vengeful, and he was a great slayer of men” (CSI, vol. 2).]8 Although Þorgeirr is described in more respectful terms, much of this description actually also fits him quite well. The farmer is a truly comic figure, faint-hearted and niggardly, although he is well off. The laws of hospitality force him to give the travellers shelter for the night and serve them food, but his lack of spirit and generosity is shown by the meal he serves: “Skammr er skutill minn,” segir Þorkell, “ok gakk þú hingat, Þorgeirr, ok sit hjá Butralda.” Þorgeirr gerir svá, gengr um þvert gólf ok sezk niðr hjá Butralda undir borðs endann. Frá verðgetum er sagt vandliga: Tveir diskar váru fram bornir; þá var eitt skammrifsstykki fornt á disk­ inum hvárum ok forn ostr til gnœttar. Butraldi signdi skamma stund,

8 Quotations in Icelandic are from the editions in the series Íslenzk fornrit (abbreviated ÍF), vols. 3 (Reykjavík 1938) and 6 (Reykjavík 1948), while quotations in English are from The Complete Sagas of Icelanders (abbreviated CSI), vols. 2 and 4 (Reykjavík 1997).

185  Scenes from Daily Life in the Sagas of Icelanders tekr upp skammrifit ok skerr ok neytir ok leggr eigi niðr, fyrr en allt var rutt af rifjum. Þorgeirr tók upp ostinn ok skar af slíkt er honum sýndisk; var hann harðr ok torsóttr. Hvárrgi þeira vildi deila við annan kníf né kjo˛tstykki. En þó at þeim væri lítt verðr vandaðr, þá fóru þeir þó eigi til sjálfir at skepja sér mat, því at þeim þótti þat sko˛mm sinnar karlmennsku (ÍF, vol. 6, pp. 144–45). [“I don’t have much to offer,” said Thorkel, “but come, Thorgeir, sit here beside Butraldi.” Thorgeir did so. He walked across the room and sat down at the table beside Butraldi. There is a detailed report of what they ate: two platters were brought in; on one of them was some old short-rib mutton and on the other a large quantity of old cheese. Butraldi made a brief sign of the cross [implying that he did not follow this custom at all], then picked up the mutton ribs, carved off the meat and continued to eat until the bones were picked clean. Thorgeir took the cheese and cut off as much as he wanted, though it was hard and difficult to pare. Neither of them would share either the knife or the food with the other. Though the meal was not good, they did not bring out their own provisions for fear that it would be seen as a sign of weakness (CSI, vol. 2, p. 341).]

The scene is repeated the next morning, only with the roles reversed: Þorgeirr grabs the meat while Butraldi tackles the old cheese. They then leave the farm, exchange insults, and Þorgeirr kills Butraldi in a picturesque way that may be either the model for the famous description of how Skarphéðinn kills Þráinn on the ice in Njáls saga, or a parody of it, depending on which saga is older and how we interpret Fóstbrœðra saga. In Njáls saga, the hero Skarphéðinn glides on the ice covering the banks of a river, jumps across the river itself where it is not frozen, glides on towards his enemies, chops the head of their leader, and then glides away on the ice; in Fóstbrœðra saga, Þorgeirr glides on hard snow down a slope and kills Butraldi. Although more exaggerated, Skarphéðinn’s feat is described in an indisputably heroic style without irony, while in Fóstbrœðra saga this scene is written in an ornate style, more frequently found in this saga than other Íslendingasögur, and characterized, i. a.,

186  Vésteinn Ólason by alliteration. This rather ‘high’ style underlines the irony of the passage through the contrast with the matter being related. It is difficult to appreciate fully the humour and ambiguity of this episode when it is read out of context, and the irony implicit in the style does not come through in translation. The climax is reached with one of Þorgeirr’s ‘heroic’ deeds, and Butraldi is described as a formidable opponent. His killing adds one more trophy to Þorgeirr’s collection. However, the context clearly shows that both ‘heroes’ are nothing but brutal thugs. Their wanderings about the barren regions of northwest Iceland, their meagre meal and their fights, can easily be seen as a parody of the wanderings of errant knights through the greenwood and their encounters with noble knights, their feasts in castles and their single combats, sometimes fought against giants. The description of the meal just given is pure comedy, and Þorgeirr’s heroic image (he certainly tries to live up to a heroic ideal) is undermined by the description of his host, of his adversary, and the meal. The conclusion is that Fóstbrœðra saga is certainly not a closed, conventional heroic tale but an ambiguous and ironic narrative that maintains a critical distance from the heroic convention. Apart from the style, the contrast between content and form, there are of course other episodes in the saga that support such an interpretation. This does not mean that the saga is pacifistic, that it is contemptuous of heroism, physical bravery or dexterity at arms as such, but it shows that such gifts of God should only be used in the service of a good cause, and the viking ideals that Þorgeirr embodies are rejected and even ridiculed. In Heiðarvíga saga (“The Saga of the Slayings on the Heath”), there is a scene describing how food is served. No less than the scene from Fóstbrœðra saga it gives us an unexpected glimpse into the daily life of Icelanders in the Middle Ages. To explain the situation for those who do not know this saga, it should be mentioned that the oldest son of the family in question was killed abroad by Icelanders from another district. The killers then perished at sea, but the family honour had to be reclaimed by exerting vengeance. Through a carefully planned series of events Barði, the second son, has created a situation that will allow him, without seeming excessively vindictive, to attack the killers’ kinsmen, people who had nothing to do with the killing of his brother.

187  Scenes from Daily Life in the Sagas of Icelanders The following scene takes place when he and a number of men that he has gathered are prepared to undertake the dangerous ride into the other district to make the attack: Nú ferr Barði heim ok fo˛runeyti hans ok er heima nótt þá. Um morgininn býr Koll-Gríss þeim do˛gurð; en þat var siðr, at lagðr var matr á borð fyrir menn, en þá váru engir diskar. Þat varð til nýnæmis, at af hurfu þrennar deildirnar fyrir þrem mo˛nnum; gekk hann ok sagði til þess Barða. “Hef þú fram borð,” segir hann, “ok rœð ekki um þat fyrir o˛ðrum mo˛nnum.” En Þuríðr mælti, at þeim sonum hennar skyldi ekki deila do˛gurð, ok kvazk hon deila mundu. Svá gerir hann, at hann hefr borð fram, borð fyrir mann, ok deilir mat á. Þuríðr gengr þá innar ok leggr sitt stykki fyrir hvern þeira brœðra, ok var þar þá yxinsbógrinn ok brytjaðr í þrennt. Tekr hann Steingrímr til orða ok mælti: “Þó er nú brytjat stórmannliga, móðir, ok ekki áttu vanða til at gefa mo˛nnum svá kappsamliga mat, ok er á þessu mikit vanstilli, ok ertu nær óvitandi vits.” Hon svarar: “Ekki er þetta furða nein, ok máttu þetta ekki undrask, fyrir því at stœrra var Hallr, bróðir yðvarr, brytj­ aðr, ok heyrða ek yðr ekki þess geta, at þat væri nein furða” (ÍF, vol. 3, pp. 276–77). [Bardi and his companions then went home to spend the night at his farm. The following morning Koll-Gris prepared them a meal. According to the custom of the time food was placed on the wooden platters before the men, as there were no dishes then. Something unusual happened: three servings, intended for three men, had disappeared. He went and reported this to Bardi, who said, “Lay the platters and say nothing of this to anyone else.” Thurid said that her sons should not be served breakfast, but that she intended to serve them. Koll-Gris brought forth platters, a platter for each man, upon which his food was served. Thurid then went in along the hall and placed a portion before each of the brothers, which turned out to be the shoulder of the ox, split into three pieces. Steingrim spoke, saying, “You’ve carved these portions generously, mother, although you’re not usually one to serve food so eagerly.

188  Vésteinn Ólason This is completely out of proportion, and you must have nearly lost your wits.” She answered, “There’s nothing strange about it, and you needn’t be surprised, as your brother Hall was carved up into larger pieces without me hearing you mention that it was anything strange” (CSI, vol. 4, p. 104).]

This passage introduces a goading-scene, a hvo˛t. The mother is here creating a situation in which she can mock her sons and remind them of the dishonour afflicted on the family, in order to strengthen their thirst for revenge. The scene serves the heroic plot of the saga by opening a dialogue between mother and sons, and it is an attempt to make the narrative more effective by first creating a puzzle, which is then solved by Þuríðr’s speech. The dialogue that follows is indeed part of a conventional heroic pattern, and the scene as a whole functions exactly as Guðrún Gjúkadóttir’s hvo˛t in the eddic lays Hamðismál and Guðrúnarhvo˛t: the mother is sharpening the will of her sons and making them angry as they are about to ride off on a journey of revenge. It is a stock-scene of heroic narrative, charging the text with emotions and slowing down the action. However, the circumstantial manner in which the preparation for the meal is described gives us a glimpse into the daily life of these people. The narrator also uses the opportunity to mark the distance in time and emphasize the historical nature of his narrative by pointing out that things were done differently at the time when the saga takes place from how they are done in his own age. The saga has a wider scope than the eddic lay, and it can include detail for which there is no room in a kviða or lay. The details of this scene are not likely to have been recorded from oral tradition: the author of the saga, who based his narrative on heroic models that he knew from tradition, must have invented the scene with its descriptive detail as well as the dialogue. The broad approach of the saga has called for innovation and attention to detail that inevitably changes the form of heroic legend and opens it up for interpretation. Compared with Guðrún’s direct and dignified approach in the eddic lay, Þuríðr’s circumstantial way of introducing her goading words has a comic ring, and an ambiguous feeling is strengthened when we are told that the following morning, when she tries to accompany her sons on their ride for revenge,

189  Scenes from Daily Life in the Sagas of Icelanders they get rid of her by making her fall into a small stream out of which she crawls and turns back home alone, bereft of the dignity that is befitting the mother of heroes. The goading scene is looked at from a distance, as it were, and it is left to the reader to determine whether it should be taken seriously or as comic relief. In this scene we find the same irony as in the previous scene from Fóstbr ðra saga: there is an ambiguity here that is much more akin to the novel than the heroic lay. The saga audience might have remembered that Þuríðr’s mother, Þorgerðr, the daughter of Egill Skallagrímsson, actually accompanied her sons on a journey of revenge for her son, and in that case no indignity is implied.9 Although both Heiðarvíga saga and Fóstbrœðra saga can present heroes in a comic light, the attitude to heroic ethics demonstrated by the two sagas is not identical. Fóstbrœðra saga pays lip service to heroic ideals while it consistently portrays the hero Þorgeirr as a comic figure, and the other hero, Þormóðr, as a man with severe faults mixed with some positive traits. Heiðarvíga saga maintains stronger ties to heroic convention, but now and again the author creates distance between the saga and the traditional heroic tale, appearing to be very well aware of the shortsightedness of heroic conduct even if he cannot help but admire it. Two more scenes from daily life illustrate this ambiguity. The first one gives a fine image of heroic splendour, while the other shows that the hero, although calm on the surface, is emotionally tense and unable to return to normal and domestic life after the carnage he has caused. In the chapter preceding the one already quoted from Heiðarvíga saga, the brothers gather their forces, and we find a scene that does not seem to advance the action, and could for that sake be cut with no loss for the plot: Nú ríðr Barði þaðan ok kemr á Bakka, þar sem Þórdís bjó, ok stóð þar hestr so˛ðlaðr, ok skjo˛ldr stóð þar hjá, ok riðu þeir heim mikinn dyn í túnit eptir ho˛rðum velli. Þar var úti karlmaðr ok kona, ok þó hon ho˛fuð hans, ok váru þau Þórdís þar ok Oddr, ok var at vanlykðum

9 Bjarni Guðnason (1993) argues that Heiðarvíga saga is younger than Laxdœla saga, where the episode in question occurs, but opinions are divided on that issue. In any case, it cannot be excluded that the tale about Þorgerðr was known from oral tradition.

190  Vésteinn Ólason no˛kkut, er hon þó ho˛fuð hans, ok hafði hon eigi þvegit lauðr ór ho˛fði honum. Ok þegar er hann sá Barða, þá sprettr hann upp ok fagnar honum hlæjandi. Barði tók vel kveðju hans ok biðr konu lúka verki sínu ok vaska honum betr. Hann lét svá gera; ok nú býsk hann ok ferr með Barða (ÍF, vol. 3, p. 273). [Bardi then rode off and when he came to the farm at Bakki, where Thordis lived, a horse was saddled and waiting, with a shield nearby. He and his following rode with a thunder of hooves into the hayfield across a hard plain. Outside were a man and woman, who proved to be Thordis and Odd. She was washing his hair and had not yet completed the job, as his head was still full of froth. As soon as he saw Bardi he sprang to his feet and greeted him with a laugh. Bardi returned his greeting and asked the woman to finish her work and wash him properly. The man allowed her to do so, then made himself ready and set off with Bardi (CSI, vol. 4, p. 102).]

There is no obvious reason for including this picture of a widow washing her steward’s hair in the narrative; it is not so˛guligt, i. e., not really a matter for a story, and it does not seem to have any narrative function other than retardation. Nevertheless, it enlivens the narrative about the gathering of forces and makes it memorable: we can see that the men joining up with Barði are no dirty gangsters, like the thugs of Fóstbrœðra saga. When Oddr is introduced into the saga, he is thus described: “Oddr … var gildr maðr fyrir sér. Ekki var hann eins kostar fégo˛fugr eða ættstórr; þó var hann frægr maðr” (ÍF, vol. 3, p. 264). [Odd … was a man of some consequence. Though he was neither wealthy nor of good family he was well-known (CSI, vol. 4, p. 98).] Odd has a strange surname: he is called Gefnar-Oddr, which connects him with the goddess Freyja and therefore characterizes him as a ladies’ man, although he is not a nobleman. It is likely that his nickname was known from the traditions about the slayings on the heath and inspired the author to create this image of daily life. It illustrates intimacy between the man and the woman, and makes him come alive as a gallant figure. Although Oddr is not a poet, his character is the same as that of the protagonists of the sagas of skalds. In the battle itself he shows his valour and some dexterity with

191  Scenes from Daily Life in the Sagas of Icelanders words when an opponent mocks him for his amorous affairs. We may ask if the adventurous and gallant ladies’ man was a traditional type or if it was formed under the influence of romance. Whatever the answer to that question is, this episode shows us how the author of Heiðarvíga saga uses a scene from daily life to transgress the limits of traditional narrative. Here this is done in a more elegant and original manner than in the previous example, and no irony is to be found in this scene. It is, however, interesting that this scene, as well as the scene from Fóstbrœðra saga already discussed, seems to contain an allusion to the world of romance, thus making the meaning of the text more complex. My last example from Heiðarvíga saga is from its final chapter. The protagonist Barði has married a woman of one of the best Icelandic families, Auðr, the daughter of Snorri the Priest. Having lived together for more than a year Barði and Auðr leave for Norway, and there the following, totally unexpected, scene takes place: Svá bar til einn morgin, er þau váru úti í skemmu bæði, at Barði vildi sofa, en hon vildi vekja hann ok tekr eitt hœgendi lítit ok kastar í andlit honum, svá sem með glensi; hann kastar braut, ok ferr svá no˛kkurum sinnum; ok eitt sinn kastar hann til hennar ok lætr fylgja ho˛ndina; hon reiðisk við ok hefir fengit einn stein ok kastar til hans. Ok um daginn eptir drykkju stendr Barði upp ok nefnir sér vátta ok segir skilit við Auði ok segir, at hann vill eigi af henni ofríki taka né o˛ðrum monnum; ekki tjár orðum við at koma, svá er þetta fast sett (ÍF, vol. 3, pp. 325). [It happened one morning that they were both out in a nearby building; Bardi wished to sleep but his wife intended to wake him. She took a small cushion and threw it in his face, as if it were a joke. He tossed it aside and this was repeated several times. Then he threw it at her and let his hand follow [that is, he hit her]. She grew angry, picked up a stone and threw it at him. That same day, after men had gathered for drinking, Bardi stood up and named witnesses and said he was divorcing Aud, on the grounds that he would not stand for her tyranny nor anyone else’s. Nothing anyone said could dissuade him, his mind was so set on this (CSI, vol. 4, p. 128).]

192  Vésteinn Ólason The fight between the couple is so real that it could have happened yesterday: an innocent and even flirtatious pillow-fight that gets out of control and ends in disaster. At this point in the story, the heroic plot is finished and the aftermath is being related. Since the hero has still not been killed, the author, according to saga conventions, has to dispose of him in some way. In the rest of the chapter, we are briefly informed of the subsequent lives of the couple: Barði journeys to Constantinople, joins the Varangian guard, earns a good reputation and falls in battle, while “Auðr var gipt o˛ðrum ríkum manni, syni Þóris hunds, er Sigurðr hét, ok eru þaðan komnir Bjarkeyingar, inir ágæztu menn” (ÍF, vol. 3, p. 325–326). [“Aud was married to another powerful man, called Sigurd, son of Thorir the Dog. The Bjarkey clan, the finest of men, is descended from them” (CSI, vol. 4, p. 129).] Barði is here portrayed as a lonely man unable to develop lasting emotional ties to other people, and the path he was forced to choose, the path of the avenger, proves to be a dead end in his personal life. The feeling created is exactly the same as when we see the lonely hero of a western movie ride towards the sunset after he has killed those who had to be killed and lost his friends and allies in the course of the action. The narrator has here taken leave of the hero and given him the heroic death that is due to him, but he also honours another convention, to name the descendants of some of the main characters, when it is mentioned that Barði’s wife Auðr was married again and had noble descendants in Norway. This is in striking contrast to her former husband, who leaves no offspring. The saga might be asking whether it is better to leave the world nothing but a great reputation or to leave fine descendants. The saga provides no answer, and the question about the inevitability and yet the futility of revenge hovers in the air. Fóstbrœðra saga mocks the hero Þorgeirr and presents the story of his life as a comedy, while Heiðarvíga saga presents Barði as a tragic hero; individual scenes as the goading by Þuríðr function as comic relief, but the overall mode of the saga is tragic, as is shown in the way it takes leave of its main hero. It is not often that Íslendingasögur give us a glimpse into a couple’s bed-chamber, as in the example above. Gísla saga Súrssonar (“The Saga of Gísli Súrsson”) is an exception, and the scene now to be discussed ends very differently from the one in Heiðarvíga saga. In addition to

193  Scenes from Daily Life in the Sagas of Icelanders two scenes in Gísla saga in which important characters are killed in their beds in the presence of their wife or sister, a more mundane, apparently trivial, scene from a bed-chamber is given in one of its chapters. Þorkell Súrsson, the brother of the heroic Gísli, has overheard a chat between Gísli’s wife Auðr and his own wife Ásgerðr, during which Auðr suggests that Ásgerðr has more love for Auðr’s brother Vésteinn than for her own husband. Ásgerðr agrees, and their words indicate that the affair was more than just a crush. The women realize that Þorkell has heard their words when he rises from his resting place and recites a stanza saying that their words will lead to the deaths of one or more people. The same evening the following scene takes place: Þorkell neytir lítt matar um kveldit ok gengr fyrstr manna at sofa. Ok er hann var kominn í rekkju, þá kemr þar Ásgerðr ok lyptir klæðum ok ætlar niðr at leggjask. Þá tók Þorkell til orða: “Ekki ætla ek þér hér at liggja náttlangt né lengra banni.” Ásgerðr mælti: “Hví hefir svá skjótt skipazk, eða hvat berr til þess?” segir Ásgerðr. Þorkell mælti: “Bæði vitu vit nú so˛kina, þótt ek hafa lengi leyndr verit, ok mun þinn hróðr ekki at meiri, þó at ek mæla berara.” Hon svarar: “Þú munt ráða verða hugleiðing þinni um þetta, en ekki mun ek lengi þœfask til hvílunnar við þik, ok um tvá kosti áttu at velja. Sá er annarr, at þú tak við mér ok lát sem ekki sé í orðit. Ella mun ek nefna mér vátta nú þegar ok segja skilit við þik, ok mun ek láta fo˛ður minn heimta mund minn ok heimanfylgju, ok mun sá kostr, at þú hafir aldri hvíluþro˛ng af mér síðan.” Þorkell þagnaði ok mælti um síðir: “Þat ræð ek, at þú ger hvárt þér líkar, en eigi mun ek banna rekkjuna náttlangt.” Hon lýsti brátt yfir því, hvárr henni þótti betri, ok ferr þegar í rekkju sína. Eigi hafa þau lengi bæði saman legit, áðr en þau semja þetta með sér, svá sem ekki hefði í orðit (ÍF, vol. 6, pp. 32–33). [Thorkel ate very little that evening and was the first to retire to bed. Once he was there, Asgerd came to him, lifted the blanket, and was about to lie down when Thorkel said, “I will not have you lying here tonight, nor for a very long time to come.” Asgerd replied, “Why this sudden change? What is the reason for this?”

194  Vésteinn Ólason “We both know what’s behind this,” said Thorkel, “though I have been kept in the dark about it for a long time. It will not help your reputation if I speak more plainly.” “You think what you will,” answered Asgerd, “but I am not going to argue with you about whether I may sleep in this bed or not. You have a choice—either you take me in and act as if nothing has happened or I will call witnesses this minute, divorce you and have my father reclaim my bride-price and my dowry. Then you wouldn’t have to worry about my taking up room in your bed ever again.” Thorkel was quiet for a while, then he said, “I advise you to do as you wish. I shall not stop you from sleeping here all night.” She soon made clear what she wanted to do, and they had not been lying together for too long before they made up as if nothing had happened (CSI, vol. 2, p. 10).]

The types of characters we meet here are well known from heroic narrative. Þorkell is the anti-hero, an indeterminate and cowardly man who does not do his duty by honouring his obligations to his kinsmen and who generally does not keep his word. Ásgerðr is a female hero, proud, determined and passionate. Obviously, she does not have the fierce pride of Brynhildr, who wants none but the best of men. Ásgerðr, however, will accept no humiliation from her husband, and she has no scruples about using her sexual power over him to ensure this. There are fine psychological nuances here. Þorkell’s lack of character is never directly mentioned, let alone condemned, but is revealed through the contrast between his brave words and his actions, as well as repeatedly through the contrast between his own behaviour and that of his brother.10 As a matter of fact, both Ásgerðr and Þorkell act according to practical unheroic considerations, but the woman shows strength 10 The ethics of Gísla saga Súrssonar, the evaluation of the characters and their acts, is a controversial subject, especially to what degree an underlying criticism of Gísli’s heroic values is inherent in the text. I have discussed this in Vésteinn Ólason 1999 and 2003; for more or less differing interpretations, see Bredsdorff 1971: 67–81, Meulengracht Sørensen 1986, and Andersson 1968, and 2006: 77–85.

195  Scenes from Daily Life in the Sagas of Icelanders and determination. This scene seems to have more in common with comic narratives of a fabliaux-type than with a heroic lay. In spite of the conventional traits of this scene, it opens the heroic form towards daily life and invites the reader to compare the saga with different kinds of texts. All the scenes that have been discussed here do this in one way or another. They do not belong to the high points in the narrative, and they would be left out in plot-summaries because they tell of events that in themselves are unworthy of telling according to convention. Nevertheless, they are an integral part of the text, belong to the web of the tale, bring the characters closer to the readers and help them to see through the text. From the finite traditional heroic tale they create one of the sagas’ many links to infinite textuality as well as to the extra-textual reality of the past. Íslendingasögur have come to us as literature, and one of the things that make them fascinating is that we can see how they are formed by conflicting and even contradictory social and textual forces. Inclusion of details from daily life and an ambiguous attitude to heroic ethics – identification and admiration conflicting with critical or ironic attitudes – as well as allusions to other contemporary kinds of narrative, cause the sagas to be seen as foreshadowing the birth of the novel. That being said, it must be repeated that the saga and the novel are two fundamentally different kinds of narrative, not historically related at all.

Bibliography Andersson, Theodore, 1968. “Some Ambiguities in Gísla saga: A Balance Sheet.” In: Bibliography of Old Norse-Icelandic Studies. Copenhagen: Munksgaard. Pp. 7–42. ———, 2006. The Growth of the Medieval Icelandic Sagas (1180–1280). Ithaca, London: Cornell University Press. Bjarni Guðnason, 1993. Túlkun Heiðarvíga sögu, Studia Islandica, vol. 50. Reykjavík: Bókmenntafræðistofnun Háskóla Íslands. Bredsdorff, Thomas, 1971. Kaos og k rlighed. En studie i isl ndingesagaens livsbillede. Copenhagen: Gyldendal (English translation, 2001: Chaos and Love: The Philoso­ phy of the Icelandic Family Sagas. Copenhagen: Museum Tusculanum). Jolles, André, 1930. Einfache Formen: Legende, Sage, Mythe, Rätsel, Spruch, Kasus, Memorabile, Märchen, Witz. Vierte unveränderte Auflage. Tübingen: Max Nie­ meyer Verlag. (First published 1930).

196  Vésteinn Ólason Ker, W. P., 1957. Epic and Romance: Essays on Medieval Literature New York: Dover Publications. (Repr. of 2nd ed. London 1908). Lönnroth Lars, 1976. Njáls Saga: a Critical Introduction. Berkeley, Los Angeles, Lon­don: University of California Press. Meulengracht Sørensen, Preben, 1986. “Murder in marital bed: an attempt at understanding a crucial scene in Gísla saga.” In: Structure and Meaning in Old Norse Literature. New Approaches to Textual Analysis and Literary Criticism. Eds. John Lindow, Lars Lönnroth, Gerd Wolfgang Weber. The Viking Collection. Studies in Northern Civilization, vol. 3. Odense: Odense University Press. Pp. 235–63. ———, 1993. Fort lling og re. Studier i isl ndingesagaerne. Aarhus: Aarhus univer­ sitetsforlag. Propp, V., 1968. The Morphology of the Folktale, transl. Laurence Scott, 2nd ed. Austin, London: University of Texas Press. (First published in Russian 1928.) Raglan, Lord, 1936. The Hero: A Study in Tradition, Myth, and Drama. London: Methuen. Simpson, John, 1973. “Scott and Old Norse Literature.” In: Scott Bicentenary Essays, ed. Alan Bell. Edinburgh, London: Scottish Academic Press. Pp. 300–313. Taylor, Archer, 1964. “The Biographical Pattern in Traditional Narrative,” Journal of the Folklore Institute, 1: 1964, pp. 114–29. Vésteinn Ólason, 1998. Dialogues with the Viking Age: Narration and Representation in the Sagas of the Icelanders, transl. Andrew Wawn. Reykjavík: Mál og menning. ———, 1999. “Gísli Súrsson – a flawless or flawed hero?“ In: Die Aktualität der Saga. Festschrift für Hans Schottmann. Herausgg. von Stig Toftgaard Andersen. Ergänzungsbände zum Reallexikon der Germanischen Altertumskunde. Her­ ausgg. von Heinrich Beck, Dieter Geuenich, Heiko Steuer. Band 21. Berlin, New York: Walter de Gruyter. Pp. 163–75. ———, 2003. “Introduction.” In: Gisli Sursson’s Saga and the Saga of the People of Eyri. translated Martin S. Regal and Judy Quinn / edited with an introduction by Vésteinn Ólason. London: Penguin. Pp. vii–xlix. ———, 2007a. “The Icelandic Saga as a Kind of Literature with Special Reference to its Representation of Reality.” In: Learning and Understanding in the Old Norse World: Essays in Honour of Margaret Clunies Ross. Ed. Judy Quinn et al. Turnhout: Brepols. Pp. 27–47. ———, 2007b. “The Fantastic Element in Fourteenth Century Íslendingasögur: A Survey,” Gripla, 17, pp. 7–21. Vries, Jan de, 1959. Heldenlied en heldensage. Utrecht, Antwerpen: Aula-boeken. (Also in English and German). Wawn, Andrew, 2000. Vikings and Victorians: Inventing the Old North in NineteenthCentury Britain. Woodbridge: D. S. Brewer.

On Some Pragmatic and Symbolic Features of Island Representation in Old Norse Literature Kristel Zilmer University of Bergen

Book 7 of the classical masterpiece The Odyssey contains a scene where the hero Odysseus (Ulysses) describes how he was shipwrecked and washed ashore on Calypso’s island, Ogygia. He thus relates: “[…] There is an isle, Ogygia, which lies far off in the sea. [245] Therein dwells the fair-tressed daughter of Atlas, guileful Calypso, a dread goddess, and with her no one either of gods or mortals hath aught to do; but me in my wretchedness did fate bring to her hearth alone, for Zeus had smitten my swift ship with his bright thunderbolt, [250] and had shattered it in the midst of the wine-dark sea.”1 Calypso kept Odysseus with her for seven years and even promised him immortality if he would stay on the island – an offer refused by Odysseus, who longed to return to his family. At last, with Zeus’ intervention, Calypso had to let Odysseus leave the island – but his adventures were far from being over. The reason for using the quote above as an introduction to the present paper lies in the fact that it illustrates fittingly the central role that islands have played in the human mind and imagination for a long time. The sequence of events concerning the numerous islands that Odysseus reached during his voyage is perhaps one of the most well-known cases; already there and then one can trace characteristic motifs developed in western thought in connection with islands. Another popular example is the medieval Navigatio Brendani, a story about the legendary voyage of Brendan of Clonfert, a sixthcentury Irish monk who is said to have travelled from one island to another in search of the Isle of the Blessed (the Promised Land of the Saints).2 1 The English translation is given according to the Perseus Digital Library text collections at www.perseus.tufts.edu/hopper. The Greek text can also be found there. 2 There are more than 120 medieval manuscripts that relate the story of Brendan; the oldest one dates from the 10th century but it is considered likely that the legend as such had achieved a full-bodied form already by the beginning of the 9th century (see e. g. Anderson 1988: 316). Approaching the Viking Age. Proceedings of the International Conference on Old Norse Literature, Mythology, Culture, Social Life and Language. Edited by Ērika Sausverde and Ieva Steponavičiūtė. (Scandinavistica Vilnensis 2). Vilnius University Press, 2009. https://doi.org/10.15388/ScandinavisticaVilnensis.2009.2.12 Copyright © 2019 Authors. Published by Vilnius University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

198  Kristel Zilmer During different times and within varying cultural contexts the themes and approaches can naturally vary a lot. But at the same time, there is ample reason to claim that there has existed and still exists a universal fascination with islands. This is what we could call the phenomenon of ‘islomania’ characterised as “[…] a central feature of Western culture, a core idea that has been a driving force from ancient times to the present” (Gillis 2004: 1). According to Gillis, such an attraction does not simply refer to being interested in islands as particular features of landscape, but also to a deeper epistemological experience: “Dividing the world into discrete things, islanding it as a means of understanding, is a peculiarly Western way of navigating a world that seems otherwise without shape and direction” (op. cit.: 2). The main focus of this paper concerns the discourse of islands on the basis of medieval Old Norse-Icelandic narrative tradition – one chapter in the history of island representation. According to our view, the imagery of islands in Old Norse sources is simultaneously reflective of certain contemporary European ideas, as well as of specific Nordic experiences that also bear witness to cultural-historical awareness.3 The emphasis in this paper is twofold: for one, we shall concentrate upon the representation of particular islands and events. Such a perspective serves at the same time to illuminate the above-mentioned Nordic dimension of island experiences. Secondly, we shall discuss the manner in which certain more general and symbolic features can be attached to the concept of island as mediated by the sources. This latter viewpoint can parallelly be applied to illustrate some facets of a broader medieval tradition on islands. Theoretical considerations In the introduction the source material was defined as medieval Old Norse-Icelandic narrative tradition. Needless to say, this paper will discuss only a limited fraction of the overall material. We have chosen to use particular forms of saga literature as the point of departure – more

3 A similar combinatory approach is also followed in the recent article “Scenes of Island Encounters in Icelandic Sagas – Reflections of Cultural Memory” by Zilmer (2008a).

199  On Some Pragmatic and Symbolic Features… precisely, the main emphasis is laid upon relevant examples deriving from the sagas of Icelanders (Íslendingasögur) and the kings’ sagas (konungasögur) – as is well known, the former focus upon the activities (conflicts) of significant Icelandic families mainly in the 10th and early 11th centuries, whereas the latter relate the history of primarily Norwegian kings through different periods of time.4 With the sagas themselves being written down some time during the 12th–14th centuries, they treat the subject matter from a retrospective and predominantly realistic point of view, making the claim of being “historical sagas about the past,” to quote Meulengracht Sørensen (1993: 98).5 In addition to discussing the island imagery in sagas, there will be a few parallels drawn to the motifs occurring in Old Norse poetry – meaning first and foremost the skaldic praise poetry of the Viking Age and the early Middle Ages. It is logical to include corresponding examples due to the very fact that they are mediated as quotations in the same sagas. It is indeed thanks to the eagerness of medieval saga writers to illustrate and document their statements in terms of (potentially authentic) skaldic stanzas that this category of Old Norse poetry has been preserved.6 But besides skaldic poetry, we shall also briefly refer to the island theme in the framework of the mythological tradition of eddic





4 The temporal frames of individual kings’ sagas are summarised e. g. by Jackson (1993: 10–15). A considerable part of the kings’ saga tradition centres around the two missionary Norwegian kings, i. e. Óláfr Tryggvason (ruled ca. 995– 1000) and Óláfr Haraldsson (ruled ca. 1015–1030). 5 Another matter is whether sagas can be analysed as historical sources as well. During the past few decades scholars have started to treat sagas as potential sources of cultural history and the history of mentality. For a discussion of different trends in recent saga scholarship, see e. g. Lönnroth (1993). 6 The authenticity and credibility of preserved skaldic poetry is another widely discussed matter. One common view is summarised by Frank (1985: 173): “Today almost none of the verse in the family sagas is considered secure; poetry in the kings’ sagas still commands credence, for it has not yet seemed likely that these verses are fabrications, falsely attributed to the early skalds.” However, the debates around individual poems continue. Concerning the  source value of skaldic poetry and the  criteria for tracing genuine stanzas, see also Bjarni Einarsson (1974); Vésteinn Ólasson (1987); Jesch (1993; 2001).

200  Kristel Zilmer poetry as this was (still) known in medieval times.7 When consulting different types of sagas as well as poetic texts, it is possible to illustrate the application of particular themes across various genres of Old Norse literature and in this manner demonstrate their common character. In further studies the scope of the source material can be enwidened to include other forms of saga literature and medieval prose tradition. The main arenas of action in Íslendingasögur and konungasögur are Iceland and Scandinavia, respectively. The latter area also lies in the focus of much of the skaldic praise poetry, whereas with regard to eddic poetry we cannot determine the precise setting in the same manner.8 Besides the main setting, sagas also include references to various other travel destinations – travel is in itself a popular motif in many a saga narrative; or as it has been put, the sagas “[…] are full of movement and experience, at home and abroad” ( Jesch 2005: 134). The image of a travelling poet, chieftain, king or simply a fame- and fortune-seeking young man is equally well recorded in Íslendingasögur and konungasögur; both types of sources cast light upon movements on a more regular scale, such as smaller raiding and trading enterprises, as well as upon larger campaigns and expeditions. Similarly, much of the skaldic poetry contains elements of travelogue in terms of listing the sites where the honoured kings and chieftains headed to. As expected, the content is then poetically modified; the act of travelling interests the skald in the framework of dramatic events and is combined with expressive battle imagery. In this current context we shall examine islands as belonging within the scheme of travelling. This allows us to draw attention to several characteristic features of the overall island representation. In order to limit the scope of study, we have chosen to concentrate mainly upon two geographical settings. For one, we look at the islands in the Baltic



7 As sources, eddic and skaldic poetry are obviously rather different with regard to their context of preservation, matter of authorship, formal and stylistic criteria as well as main content. For an overview of the characteristic features of both types of poetry, see e. g. Hallberg (2003), Mundal (2004). Certain parallels between skaldic and eddic poetry are discussed by Zilmer (2008b). 8 However, as will be shown during the analysis, among the toponyms inserted into the eddic poetry there do occur a few identifiable island references as well.

201  On Some Pragmatic and Symbolic Features… Sea, in the region of the Kattegat and the Danish straits – an active arena for communication. The second setting is made up by islands along the western coast of Norway that are depicted to host frequent traffic in the context of western travels along the North and the Norwegian Seas.9 Both of the above-mentioned territories demonstrate clear communicative significance as attested to by the sagas. A final consideration concerns the definition of an island in this study. Determining what exactly an island is can be a tricky linguisticsemantic, cultural-historical as well as geographical matter. On the one hand, one could analyse the application of particular linguistic elements in Old Norse place names and look at various descriptive references provided by the sources.10 On the other hand, it would be necessary to take into consideration different contextual factors and among other things to discuss the relationship between island(s) and mainland(s). Certain puzzlement concerning the distinction between islands and mainlands is visible from the sources as well. In the 13th‑century Norwegian prose work known as Konungs skuggsjá (The King’s Mirror) – composed as a dialogue between a father and a son – a question is posed as to whether Greenland should be considered a mainland or an island: “Svá forvitnar mik ok þat, hvárt þér tlit at þat sé meginland eða eyland” (Keyser et al 1848: 42).11 The father then explains that the size of Greenland is unknown, but it is taken to be a mainland and that it is connected to some other mainland; the proof is found in the fact that Greenland has such animals that do not usually live on islands. That one had to get a sense of a territory in order to determine whether this was an island also shines through in a short passage in Grœnlendinga saga (ch. 2) describing the voyage of Bjarni Herjólfsson. Bjarni and his men intended to

9 Again, in further research, it is without doubt important to include other maritime territories as well; for example, there occur interesting island motifs with regard to the Northern Atlantic setting. 10 An interesting semantic matter would be to look at different terms applied in connection with islands and study specific compounds including the element ey (‘island’). See e. g. Cleasby et al (1957: 134) for examples that refer to different types of islands (inhabited ones, ones that lie far out in the sea, etc.). 11 Note the application of the word eyland (island, or literally: ‘island-land’); the same word is reflected in the modern Swedish proper name Öland designating an island off the eastern coast of Sweden.

202  Kristel Zilmer sail to Greenland, but due to unfavourable sailing conditions came across several unknown lands. Approaching a third unknown territory they: “[…] halda með landinu fram ok sá, at þat var eyland” (Matthías Þórðarson and Einar Ólafur Sveinsson 1935: 247).12 A few general concepts to keep in mind in the meantime are the following. In Old Norse cosmography (and similarly in ancient geography), we find the notion of the inhabited world being an island, surrounded by the chaotic Ocean, in Old Norse known as Úthaf (see e. g. Hastrup 1990: 28–29; Gillis 2004: 12–13). Considering the model that sees the whole world as an island, it seems obvious that the size (and the status) of the landmass is not necessarily an apparent feature to pursue when defining (is)lands. Also, an island does not even have to be surrounded by water; according to Gillis (2004: 17), in medieval times any strange and distant place could be viewed as an island.13 Here we shall follow a somewhat simplified practice and still define islands first and foremost as waterbound insular communities, emerging as such on the basis of the narrative context itself. This approach is motivated by the actual perspective of the sources; as is for example said in the 13th-century work known as Snorra Edda or The Younger Edda (according to the manuscript Codex Upsaliensis, DG 11): “[…] en ey heitir þat land sem sior eða vatn fellr vm hverfis” ( Jón Sigurðsson 1852: 366).14 Islands as significant sites The treatment of islands in Íslendingasögur and konungasögur as well as in skaldic poetry in many ways presents them as useful sites for localising various events. Furthermore, they emerge as important cognitive landmarks from the maritime perspective, bearing witness to the sailing experiences of the Northmen. On the whole, we can say that the role of islands in the narrative is very much connected to the typical activities of a travelling hero. One distinctive facet of island imagery in the sagas and in skaldic poetry is that they provide a suitable setting for battles, campaigns and raids. The fighting may then occur either directly on 12 ‘[…] [they] followed the land and saw that it was an island’ (my translation). 13 Gillis (2004: 61–64) also explains that it was only in the 16th century that one started to make a clearer distinction between islands and continents. 14 ‘[…] an island is a land which is surrounded by sea or lake’ (my translation).

203  On Some Pragmatic and Symbolic Features… the island or in the waters around it; various features of the landscape may be illuminated, showing for example how islands could be used to organise stakeouts, etc. Corresponding motifs are well-recorded in konungasögur, but also meet us in certain Íslendingasögur. At one extreme, such island confrontations include big battles between Scandinavian kings; at the other, private duels between two opponents. In fact, the concept of duelling creates an association to islands also in terms of its name, i. e. hólmganga, which contains a reference to small islands (holmr / hólmr). As we learn from the sagas, an alternative to the island setting is to carry out a duel on top of a small hillock; in the meantime, even such a site appears as a kind of island in relation to the surrounding landscape. But the sagas do also refer to actual island duels – and those may involve fighting a vicious warrior or even a supernatural creature.15 We shall illustrate the motif of islands as key (battle) sites by taking a look at the island of Hlésey (Læsø) in the middle of the Kattegat. In the saga on Magnús blindi and Haraldr gilli in the 13th-century Norwegian kings’ saga compilation Heimskringla, we hear about a battle Haraldr held by that island (see ch. 12 of Magnúss saga blinda ok Haralds gilla). The same event is referred to by the somewhat earlier kings’ saga compilations Morkinskinna and Fagrskinna. In all three cases the saga statements are illustrated by a skaldic quote assigned to a 12th‑century poet Einarr Skúlason; here the whole stanza is given according to the skaldic poetry edition: “O ˛ ttuð sókn við sléttan, / serkrjóðr Ho˛ars, merki, / harðr, þars hregg of virðum, / Hléseyjar þro˛m, blésu; / hús brann up, en eisur / ófatt, séa knátti, / malmr so˛ng, en hlóð hilmir / hr ko˛st, við ský gn fa” (Finnur Jónsson 1912–1915 B I: 424–425).16 15 See e. g. ch. 65 of Egils saga Skalla-Grímssonar that contains a scene of Egill fighting with a berserk, or ch. 18 of Grettis saga Ásmundarsonar where Grettir fights against a troublesome mound-dweller. 16 “The hardened colorer of Hálfr’s [legendary king] shirt [byrnie, warrior] held battle by the level shore of Hlésey [Læsø] where the storm caused standards to billow above the men. Many a house was consumed by fire, and one could see flames leaping against the clouds; steel sang, and the king stacked a corpse pile.” Translation according to the English version of Morkinskinna (Andersson and Gade 2000: 365). Note that Morkinskinna quotes the whole stanza whereas Fagrskinna and Heimskringla give only the first half of it.

204  Kristel Zilmer Before mentioning the battle at Læsø, the sagas point out another island and again an illustrative skaldic quote is inserted into the narrative. This island is identified as Hveðn (Ven), a small island in the strait of Øresund, between Sjælland and Skåne. Fagrskinna also specifies here that Haraldr was fighting against some Vikings. In this context islands are thus identified as fitting sites for holding victorious battles. However, island encounters do not always prove to be successful for the saga character; the narratives also highlight potential dangers connected to such places. In ch. 94 of Óláfs saga helga in Heimskringla, we hear about a certain Gauti Tófason; his story is told as news to the Swedish king by a wise man called Emundr from Skara.17 One time Gauti Tófason sailed along the Göta River and when he reached Eikreyjar (referring to the northern Göteborg archipelago where we also find the island of Öckerö) he noticed five Danish trade ships. Gauti conquered four of them together with his men, and then started chasing the last one with one of his ships, but lost sight of it. Due to a heavy storm he lost his ship and all the men on board – this happened by the island of Læsø. Meanwhile, his remaining companions who were still waiting for him by Eikreyjar themselves got attacked by more Danes, and now the fortune turned, because they all got killed. Besides its informative value this little tale carries allegoric significance, since it demonstrates how one loses everything as a result of having got too greedy. In this manner, Emundr also prepares the ground for bringing up a more important case of complaint in front of the king. This latter example referred to Læsø in the context of stormy weather leading to a shipwreck. Islands can thus figure in the scheme of events that highlight the hardships caused by weather and problematic sailing conditions – in a situation like that, battling becomes much more challenging. Læsø is also mentioned in ch. 35 of Haralds saga Sigurðarsonar in Heimskringla. King Haraldr, who has been raiding around in Denmark, is sailing northwards in the Kattegat when contrary winds force him to take lee by Læsø. The same scene is brought up by Fagrskinna (see ch. 55) and Morkinskinna (p. 166). In the meantime, the latter two accounts claim that Haraldr had to find shelter by 17 The same scene is analysed in Zilmer (2005: 286).

205  On Some Pragmatic and Symbolic Features… a different Danish island – namely Sámsey (Samsø). In all three cases there is also mention of a heavy fog occurring on the island that forms a contrast to the sun reflecting from the ships out on the sea – those being the approaching ships of Haraldr’s opponent, the Danish king Sveinn Úlfsson. As explained by Finlay (2004: 206), the reference to Læsø in Heimskringla can be considered “more probable in relation to Limafjo˛rðr” (i. e. the sound of Limfjord in Jutland), also forming part of the setting. On the other hand, Samsø gets mentioned as an important locality earlier in Morkinskinna in connection with King Magnús góði; the island is then characterised as the usual anchorage for the king: “[…] þeir kuomu uid Samsey og lagu skipunum þar j einre hofn sem jafnan hafde legit Magnus konungr fyrr” (Finnur Jónsson 1932b: 146).18 It is not always necessary to identify the location precisely; more important is to show that the site as such fulfils its particular narrative purpose. On certain occasions, it can also be expected that specific contextual clues would make it clear which place one had in mind. Among famous battle sites, the island of Svo˛lðr (i. e. Svolder) is given as the setting for the final battle of Óláfr Tryggvason.19 As for example stated in Fagrskinna (ch. 24), the enemies of the king had gathered their forces there: “Við einn hólma fyrir Vinðlandi váru saman komnir margir stórir ho˛fðingjar. Þessi hólmi heitir Svo˛lðr” (Bjarni Einarsson 1985: 147).20 In the preserved skaldic poetry relating of that battle, the name of the island is not necessarily specified. Erfidrápa Óláfs Tryggvasonar by Hallfreðr Óttarsson vandræðaskáld, for example, localises the battle in the south of the sea; the location is further described as the broad sound of an island (á víðu Holms sundi, see stanza 17). The southern direction is also emphasised in Eiríksflokkr by Haldórr ókristni; and there as well a 18 “[…] they came to Sámsey (Samsø) and anchored in the harbor where King Magnús had always anchored off that island in earlier days” (Andersson and Gade 2000: 185). 19 The exact location of Svolder is unknown; it has for example been suggested that it is an island somewhere in the southern Baltic close to the German coast. See the short overview provided by Andersson in his translation of The Saga of Olaf Tryggvason by Oddr Snorrason (2003: 147); cf. also Finlay (2004: 116). 20 “By an island off the coast of Vinðland there were gathered many important chieftains. This island is called Svo˛lðr” (Finlay 2004: 116). See also ch. 99 of Óláfs saga Tryggvasonar in Heimskringla.

206  Kristel Zilmer reference is provided to the island in terms of applying the word holmr (see stanza 3). Perhaps in these cases it is the southern setting itself that makes the locality clear, so that no further identification is required.21 On the other hand, sometimes the sagas may find it important to identify even small and uninhabited islands and provide descriptive comments as to their particular features. One such example concerns Brenneyjar (in modern Swedish Brännöarna) – another archipelago on the western coast of Sweden. These small islands are mentioned on several occasions; and in Bjarnar saga (ch. 7) it is explained: “[…] þat eru margar eyjar ok váru þá lítt byggðar. Þar váru í launvágar, ok var þar jafnan herskátt af víkingum; skógr var þar ok no˛kkurr á eyjunni” (Guðni Jónsson and Sigurður Nordal 1938: 127).22 Similarly in ch. 48 of Egils saga Skalla-Grímssonar, we hear that Vikings used to wait for passing trade ships in the region of Brenneyjar. The examples above have illustrated certain islands as the focal points in the southern Baltic. Other islands in the same region (but also in other maritime landscapes) fulfil similar functions. In the case of the bigger Danish islands, such as Sjælland, Fyn, and Falster, it is obvious that besides being common battle sites they also belong with the picture of extending one’s territorial dominion – much traffic is taking place in between these and other islands as a result of the political confrontations between various Danish and Norwegian rulers. But in addition to that, the depiction of islands in the Baltic Sea also expresses strategic navigational purposes. The islands form part of common sailing routes, and the knowledge connected to them has pragmatical significance.23 21 At the same time the name Svo˛lðr is recorded in a poem by Skúli Þórsteinsson; the skald speaks of a battle sunnr fyr Svo˛lðrar mynni i. e. ‘south by the mouth of Svolder’ (see stanza 2) which according to Finlay (2004: 116) may create the impression of Svo˛lðr being a river. However, sagas relate that Svo˛lðr was an island. The phrase “the mouth of Svolder” could perhaps be taken as referring to a sound connected to that island. Furthermore, certain formulations in skaldic poetry are only motivated by stylistic considerations. 22 “This is a group of many islands, not much inhabited at that time. There were hidden creeks in them, and they were always exposed to raiders. There was also some woodland on the islands” (Finlay 1997: 263). 23 The strategic significance of islands in the Baltic sea as depicted in various Old Norse sources is also emphasised in Zilmer (2006, see pp. 259–267).

207  On Some Pragmatic and Symbolic Features… On the one hand, islands may appear as navigational aids, suitable outposts and anchorages; on the other hand, it is important to stay aware of the challenges that sailing around certain islands could pose. Such are the very practical aspects of island representation that can be traced on the basis of both narrative and directly geographical sources. In the following, we shall concentrate upon a more symbolic facet of island representation. Earlier we referred to such island combats where one has to fight against some remarkable or even supernatural creature. A quote from eddic poetry that refers to the island of Læsø from a similar perspective can provide a fitting transition from the practical to the symbolic. Thus, in the poem Hárbarðsljóð, stanza 37, we hear about the god Þórr fighting against some berserk women on Læsø – combining a clear identification of a well-known maritime site with mythological motifs. This is what Þórr says about his island experience: “Brúðir bersęrkja / barðak í Hlésęyju, / þ r ho˛fðu verst unnit, / vélta þjóð alla” (Finnur Jónsson 1932a: 87).24 Island symbolics This brings us to the symbolic level of island representation, which in many ways accords with a broader medieval tradition on islands. Islands thus get connected with peculiar events and appear as mysterious and miraculous sites – both in the positive and negative sense. To start with the latter, in sagas we for example meet the motif of islands as sites for outlaws and criminals. Due to its relative isolation, an island can provide a perfect hiding place and / or prevent contact. Similar perceptions shine through in an island reference occurring among eddic poems, namely in Vo˛lundarkviða (see Finnur Jónsson 1932a: 125–126). A small fictional island called S varsto˛ð (meaning ‘sea-harbour’) figures as the place where the smith Vo˛lundr is kept in captivity and has to work for the king. No one else is supposed to have contact with him; this does not really work out, and later in the poem we hear about some horrific incidents taking place on the island. 24 “Berserk women I fought in Hlesey, / they’d done the worst things, bewitched all men” (Larrington 1996: 74).

208  Kristel Zilmer Another motif concerns the somewhat strange habits of people living on islands, as well as the fact that islands may provide a home base for weird and dangerous creatures. With regard to the former, Old Norse even had a special term coined for marking islanders – they could be called eyjar skeggjar, i. e. ‘island beards’ (see Cleasby et al 1957: 134). Concerning the latter aspect, we already referred to certain motifs in sagas (for example, Grettis saga Ásmundarsonar) that speak about encounters with island monsters. In accordance with common medieval perceptions, an island itself may appear as a monster or a gigantic creature of some kind. The abovementioned Navigatio Brendani includes a scene concerning the so-called Fish Island ( Jasconius). Brendan and his followers camp on the back of the giant fish, believing it to be an island; they even celebrate Easter there. Once they light a fire on its back, the monster awakes and starts moving and the men have to flee. Parallels to that particular story and / or other similar motifs can perhaps be found in a little humouristic episode related in ch. 3 of Knýtlinga saga and ch. 33 of Óláfs saga Tryggvasonar in Heimskringla. In this connection, we hear about problems the Danish king Haraldr Gormsson was having with Icelanders who were producing mocking verses of him. One of the king’s sorcerers is sent on a magic ride to Iceland in the form of a whale; remarkably enough, he later informs the king that there live all kinds of monstrous creatures on that land, and that it also lies way too far to be reached by ship. This little comment provides a kind of side-step regarding the image of Iceland, in supposedly representing the island’s role in the eyes of those from the outside, and fitting well with the medieval practice of depicting remote and mysterious places. But islands could also be given religious significance; according to another medieval concept islands emerge as holy and sacred sites (see Gillis 2004: 26–39). In Old Norse sources islands may be associated with both heathen and Christian practices; on more specific occasions there are even magic activities taking place there. Looking at an example from eddic poetry, in stanza 24 of Lokasenna Loki accuses Óðinn of having engaged in disgraceful magic practices (seiðr) on the island of Samsø: “Ęn þik síða / kóðu Sámsęyju í, / ok drapt á vett sem vo˛lur, / vitka

209  On Some Pragmatic and Symbolic Features… líki / fórt verþjóð yfir, / ok hugðak þat args aðal” (Finnur Jónsson 1932a: 105).25 Again, it is interesting to observe how the very real island is connected to certain fantastic events in the framework of eddic poetry. In this current context we shall in the meantime concentrate upon the motif of islands as sites for religious transformations resulting from the act of Christianisation. The islands that are in focus are those located along the western coast of Norway connected to the missionary activities of Óláfr Tryggvason and Óláfr Haraldsson. We shall use the former as our point of departure and look at a few scenes occurring in a separate saga on Óláfs saga Tryggvasonar by Oddr Snorrason (from the end of the 12th century), as well as in the kings’ saga compilations Fagrskinna and Heimskringla.26 Óláfr is said to have sailed from the Orkney Islands to Norway after having baptised the Orkney jarl. In Norway he lands on the island of Moster (in Sunnhordland); during the night he is visited by a saint who says the king should call people to give up their old customs and accept the new religion. We further hear about Óláfr organising an assembly on the island and speaking to the people while standing on a high rock. Three men attempt to come with arguments that would oppose Óláfr, but neither among them is able to deliver his speech: “Nu firir þui at þeir voro sigraðir með sua miclum krapti. Þa tocu margir við tru. Oc firir letu forna villu oc fylgþu allir konungs boði” (Finnur Jónsson 1932c: 96).27 In the following, the saga relates about the island of Selja (in Sogn og Fjordane), which emerges as a sacred site. Several people are said to have experienced special light and sweet fragrance coming from the island, and they relate these stories to the king. Óláfr then heads to the island together with his bishop, and they find there a place with 25 “But you once practised seid on Samsey / and you beat on the drum as witches do, / in the likeness of a wizard you journeyed among / mankind, / and that I thought the hallmark of a pervert” (Larrington 1996: 89). Concerning the practice of seiðr, see e. g. the dissertation by Heide (2006). 26 Originally Oddr Snorrason wrote his work in Latin, but what is preserved is the Old Norse translation of it. Oddr himself was most likely influenced by both Latin hagiography and Old Norse learned texts. 27 “Because the three of them were overcome by such power; many accepted the faith and abandoned their former superstition and followed the king’s command” (Andersson 2003: 76).

210  Kristel Zilmer bones, i. e. holy relics. The king understands the religious significance of the island and thus has a church built there: “Oc at bỏn byskups oc konungs raði. var þar kirkia ger. oc helguð þessum guðs monnum er þar voro” (Finnur Jónsson 1932c: 100).28 The saga also provides a comment with regard to the island of Kinn: “Slict hit sama verþa oc morg tocn iannarri eyio er KiN heitir. ero þar oc helgir domar þessarrar sueitar. sem iSelio. Oc ueitir guð firir huarratueggio sakir margar iartegnir firir milldi sina oc miscunn” (ibid.).29 Regarding the religious community of Selja, the saga includes a story concerning the tradition on Saint Sunnifa. According to the legend, she was an Irish princess who abandoned her home to escape marrying a heathen man. By the good will of God, she and her people came to the island of Selja. Selja also became the death place of Sunnifa and her followers – the caves on the island collapsed over them, saving them from falling into the hands of local heathens who thought they were causing trouble on the island. The saga also remarks that there are many miracles connected with the island and provides the names of the churches that have been built there.30 As a parallel to the discussed episodes in Óláfs saga Tryggvasonar by Oddr Snorrason, we can mention that ch. 47 of Óláfs saga Tryggvasonar in Heimskringla contains a short reference to the island of Moster. The island is similarly identified as the site where Óláfr Tryggvason first came ashore. According to Heimskringla, a mass was sung on the island, and later a church was built there. Heimskringla does not relate of Selja in connection with Óláfr Tryggvason. In the meantime, Selja is mentioned in ch. 29 of Óláfs saga helga – with certain parallels noticeable between this reference to the activities of Óláfr Haraldsson and the one that relates of Óláfr Tryggvason’s arrival on Moster. Both islands function as outposts for the kings arriving in Norway. In the case of Óláfs saga helga, it is told 28 “At the request on the bishop and the command of the king a church was built there and dedicated to the men of God who were there” (Andersson 2003: 77). 29 “There are also many signs on another island called Kinn. There are also holy relics of the same community as on Selja. In honor of both, God performs many miracles in his graciousness and mercy” (Andersson 2003: 77). 30 A detailed account on the tradition concerning Sunnifa and the actual finds made on the island of Selja is given in Rindal (ed., 1997).

211  On Some Pragmatic and Symbolic Features… that during the return trip to Norway Ólafr and his men experienced trouble on the sea, but thanks to the good luck of the king everything went well. They landed on the island, which in Heimskringla is given the name S la, meaning luck and happiness. This is naturally taken as a good sign by the king. Fagrskinna contains a similar scene in connection with Óláfr Haraldsson (see ch. 28): “Þeir kómu at hafi útan at Staði ok þar á land, sem ey ein lítil er ok heitir S la. Þá m lti Óláfr ok lét þá tímadag hafa land tekit ok talði þat gott mark, at þeir váru komnir í S lu” (Bjarni Einarsson 1985: 170).31 In both Heimskringla and Fagrskinna, it is further related how the king stumbled on land, an unlucky accident. In the meantime, his companion tried to save the situation by stating that in this manner the king established his power over Norway. The symbolics connected to the island of Selja as it is represented in saga narratives, as well as other sources that relate of Sunnifa, demonstrates that the site was considered important for several reasons. The explicit saga motifs show Selja as a holy site, with clear religious significance. From a comparative perspective, it has been pointed out that one can find elucidative parallels between Selja and certain holy sites along the British coasts. As argued by Crawford (1997: 178), it must have been necessary for the kings to find a suitable locality for establishing a religious cult; and according to the Celtic tradition, an island off the coast was especially well suited for such purposes. In this way, we can also notice missionary and political motivation behind the scenes that are reflected in the sagas. Also, the emphasis upon Selja and other similar islands is not accidental. As has been pointed out for example by Hommedal (1997: 63–65), the location of Selja is indeed strategic, and the island made up a perfect natural harbour for people travelling past the Stad peninsula. The latter promontory is known because of complicated weather conditions that can occur in its neighbourhood. Thus, for travellers Selja must have made up an ideal anchorage when they had to wait out harsh weather, for 31 “They came in from the sea at Staðr, and came to land where there is a little island called Sæla. Then Óláfr spoke, saying that the day of their reaching land was a lucky one, and he reckoned it a good sign that they had come to Sæla (Happiness)” (Finlay 2004: 137). As commented by Finlay (ibid.), “S la has been considered a variant of Selja, adopted for the sake of the pun”.

212  Kristel Zilmer example.32 In this way the circle is complete, since we have returned from symbolic and religious features to the visibly pragmatical aspects of island representation. Conclusions Island representation in Old Norse narrative sources is based upon insights from the practical activities and experiences of the Northmen as well as the more general medieval concepts concerning communication with islands. In this current paper we have concentrated upon a few limited aspects of such island imagery. Islands – both real and imaginary, named and unnamed – appear in different types of narrative contexts and fulfil a variety of functions. On the one hand, islands may emerge as significant (battle) sites that are further characterised by their strategic positioning on the maritime landscape; on the other hand, it has been shown that there are special religious and mythological motifs attached to them, which illuminate their peculiarity. The examples discussed above have at the same time demonstrated that one dimension does not have to exclude the other – we can rather witness a combination of various types of features across the sources. This concerns both the imagery of particular islands (take for example the Norwegian island of Selja), as well as the broader understandings on islands as spatial and cognitive figures. In this we can witness a mixture of various narrative depiction techniques, as well as the blending of actual cultural traditions. All in all, it is obvious that the Old Norse sources in their own way bring out the many dynamic qualities of island representation in the human imagination. As such, there are various ways in which the mediated images can contribute to our ongoing fascination with islands.

32 Similar perspectives can be associated with other islands as well. The motif of the island of Moster functioning as a site where ships could be stationed is for example included in ch. 31 of Haraldssona saga in Heimskringla.

213  On Some Pragmatic and Symbolic Features… Bibliography Andersson, Theodore M., trans. 2003. [Oddr Snorrason] The Saga of Olaf Tryggvason. Islandica 52. Ithaca and London: Cornell University Press. Andersson, Theodore M. and Kari Ellen Gade, trans. 2000. Morkinskinna. The Earliest Icelandic Chronicle of the Norwegian Kings (1030–1157), Islandica 51. Ithaca and London: Cornell University Press. Anderson, John D. 1988. “The  Navigatio Brendani: A Medieval Best Seller”, in The Classical Journal Vol. 83, No. 4. 315–322. Bjarni Aðalbjarnarson, ed. 1941–1951. [Snorri Sturluson] Heimskringla, Íslenzk fornrit 26–28. Reykjavík: Hið íslenzka fornritafélag. Bjarni Einarsson. 1974. “On the Role of Verse in Saga-Literature”, in Mediaeval Scandinavia 7. 118–125. Bjarni Einarsson, ed. 1985. Fagrskinna, Íslenzk fornrit 29. Reykjavík: Hið íslenzka fornritafélag. 55–373. Bjarni Guðnason, ed. 1982. Knýtlinga Saga, Íslenzk fornrit 35. Reykjavík: Hið íslenzka fornritafélag. 91–321. Cleasby, Richard, Gudbrand Vigfusson and W. A. Craigie. 1957. An Icelandic-English Dictionary. Oxford: Clarendon. Crawford, Barbara E. 1997. “Hellige steder på De britiske øyene. Noen paralleller til Selja”, in Magnus Rindal, ed. Selja – heilag stad i 1000 år. Oslo: Universitetsfor­ laget. 160–182. Finlay, Alison, trans. 1997. The Saga of Bjorn, Champion of the Hitardal People, The Com­plete Sagas of Icelanders 1. Reykjavík: Leifur Eiríksson Publishing. 255–304. Finlay, Alison, trans. 2004. Fagrskinna, a Catalogue of the Kings of Norway, The North­ ern World, Vol. 7. Leiden and Boston: Brill. Finnur Jónsson, ed. 1912–1915. Den norsk-islandske skjaldedigtning. Copenhagen: Villadsen og Christensen. Finnur Jónsson, ed. 1932a. De gamle eddadigte. Copenhagen: Gads forlag. Finnur Jónsson, ed. 1932b. Morkinskinna. Copenhagen: J. Jørgensen & Co. Finnur Jónsson, ed. 1932c. Saga Óláfs Tryggvasonar af Oddr Snorrason Munk. Copen­ hagen: Gads forlag. Frank, Roberta. 1985. “Skaldic Poetry”, in Carol J. Clover & John Lindow, eds. Old Norse-Icelandic Literature. A Critical Guide, Islandica 45. Ithaca & London: Cornell University Press. 157–196. Gillis, John. 2004. Islands of the Mind. How the Human Imagination Created the Atlan­ tic World. New York: Palgrave Macmillan. Guðni Jónsson, ed. 1936. Grettis saga Ásmundarsonar, Íslenzk fornrit 7. Reykjavík: Hið íslenzka fornritafélag. Guðni Jónsson and Sigurður Nordal, eds. 1938. Bjarnar saga hítdœlakappa, Íslenzk fornrit 3. Reykjavík: Hið íslenzka fornritafélag. 109–211.

214  Kristel Zilmer Hallberg, Peter. 2003 [1964]. Den fornisländska poesien, Verdandis skriftserie 20. Stockholm: Svenska bokförlaget (Bonnier). Hastrup, Kirsten. 1990. Island of Anthropology. Studies in Past and Present Iceland. Odense: Odense University Press. Heide, Eldar. 2006. Gand, seid og åndevind. Bergen: University of Bergen. Homer. Odyssey. Perseus Digital Library. www.perseus.tufts.edu/hopper [accessed 18.04.2008]. Hommedal, Alf Tore. 1997. “Bakgrunnen for helgenanlegget på Selja og staden si rolle i den tidlege kristninga av Vest-Noreg”, in Magnus Rindal, ed. Selja – heilag stad i 1000 år. Oslo: Universitetsforlaget. 43–76. Jackson, Tatjana = Джаксон, Т. Н. 1993. Исландские королевские саги о Восточной Европе (с древнейших времен до 1000 г.): тексты, перевод, комментарий). Москва: Наука. Jesch, Judith. 1993. “Skaldic Verse and Viking Semantics”, in Anthony Faulkes & Richard Perkins, eds. Viking Revaluations. Viking Society centenary symposium, 14–15 May 1992. London: Viking Society for Northern Research. 160–171. Jesch, Judith. 2001. Ships and Men in the Late Viking Age. The Vocabulary of Runic Inscriptions and Skaldic Verse. Woodbridge: The Boydell Press. Jesch, Judith. 2005. “Geography and Travel”, in Rory McTurk, ed. A Companion to Old Norse-Icelandic Literature and Culture. UK: Blackwell Publishing. Jón Sigurðsson, ed. 1852. Edda Snorra Sturlusonar II. Edda Snorronis Sturl i II. Hafniæ: Legatus Arnamagnæani. Keyser, Rudolph, Peter Andreas Munch, Carl Rikard Unger, eds. 1848. Speculum Regale. Konungs-skuggsjá. Konge-speilet. Et philosophisk-didaktisk skrift, forfattet i Norge mod slutningen af det tolfte aarhundrede. Christiania: Carl C. Werner & Comp. Larrington, Carolyne, trans. 1996. The Poetic Edda. Oxford: Oxford University Press. Lönnroth, Lars. 1993. “På andra sidan Fyrisvallarna. Några tendenser i modern sagaforskning”, in Göran Dahlbäck, ed. Snorre Sturlasson och de isländska källorna till Sveriges historia. Fyra föreläsningar från ett symposium i Stockholm hösten 1988. Sällskapet Runica et Mediaevalia. Stockholm: Norstedt. 85–103. Matthías Þórðarson and Einar Ólafur Sveinsson, eds. 1935. Grœnlendinga saga, Íslenzk fornrit 4. Reykjavík: Hið íslenzka fornritafélag. 239–269. Meulengracht Sørensen, Preben. 1993. Saga and Society. An Introduction to Old Norse Literature. Translated by John Tucker. Studia Borelia / Nordic Studies I. Odense: Odense University Press. Mundal, Else. 2004. “Edda- og skaldediktning”, in Odd Einar Haugen, ed. Handbok i norrøn filologi. Oslo: Fagbokforlaget. 215–266. Rindal, Magnus, ed. 1997. Selja – heilag stad i 1000 år. Oslo: Universitetsforlaget. Sigurður Nordal, ed. 1933. Egils saga Skalla-Grímssonar, Íslenzk fornrit 2. Reykjavík: Hið íslenzka fornritafélag.

215  On Some Pragmatic and Symbolic Features… Vésteinn Ólason 1987. “Norrøn litteratur som historisk kildemateriale”, in Gunnar Karlsson, ed. Kilderne til den tidlige middelalders historie. Rapporter til den XX nordiske historikerkongres. Reykjavík: Sagnfræðistofnun Háskóla Íslands. 30–47. Zilmer, Kristel. 2005. ‘He Drowned in Holmr’s Sea – His Cargo-ship Drifted to the Sea‑bottom, only Three Came out Alive.’: Records and Representations of Baltic Traffic in the Viking Age and the Early Middle Ages in Early Nordic Sources, Dis­ sertationes Philologiae Scandinavicae Universitatis Tartuensis 1, Nordistica Tartuensia 12, Tartu: Tartu University Press. Zilmer, Kristel. 2006. “The Representation of Waterborne Traffic in Old Norse Nar­ ratives – The Case of the Baltic Sea Area”, in Viking and Medieval Scandinavia 2. 239–274. Zilmer, Kristel. 2008a. “Scenes of Island Encounters in Icelandic Sagas – Reflections of Cultural Memory”, in Viking and Medieval Scandinavia 4. 227–248. Zilmer, Kristel. 2008b. “Vanapõhja luulekunst”, in Akadeemia 2008, no. 4. 731–768.